all | dynamic | editing | fast | generalization | human | video | lighting | reconstruction | texture | semantic | pose-slam | others
- Neural Radiance Fields from Sparse RGB-D Images for High-Quality View Synthesis, TPAMI2022 | [code]
The recently proposed neural radiance fields (NeRF) use a continuous function formulated as a multi-layer perceptron (MLP) to model the appearance and geometry of a 3D scene. This enables realistic synthesis of novel views, even for scenes with view dependent appearance. Many follow-up works have since extended NeRFs in different ways. However, a fundamental restriction of the method remains that it requires a large number of images captured from densely placed viewpoints for high-quality synthesis and the quality of the results quickly degrades when the number of captured views is insufficient. To address this problem, we propose a novel NeRF-based framework capable of high-quality view synthesis using only a sparse set of RGB-D images, which can be easily captured using cameras and LiDAR sensors on current consumer devices. First, a geometric proxy of the scene is reconstructed from the captured RGB-D images. Renderings of the reconstructed scene along with precise camera parameters can then be used to pre-train a network. Finally, the network is fine-tuned with a small number of real captured images. We further introduce a patch discriminator to supervise the network under novel views during fine-tuning, as well as a 3D color prior to improve synthesis quality. We demonstrate that our method can generate arbitrary novel views of a 3D scene from as few as 6 RGB-D images. Extensive experiments show the improvements of our method compared with the existing NeRF-based methods, including approaches that also aim to reduce the number of input images.
- Diffusion Guided Domain Adaptation of Image Generators | [code]
Can a text-to-image diffusion model be used as a training objective for adapting a GAN generator to another domain? In this paper, we show that the classifier-free guidance can be leveraged as a critic and enable generators to distill knowledge from large-scale text-to-image diffusion models. Generators can be efficiently shifted into new domains indicated by text prompts without access to groundtruth samples from target domains. We demonstrate the effectiveness and controllability of our method through extensive experiments. Although not trained to minimize CLIP loss, our model achieves equally high CLIP scores and significantly lower FID than prior work on short prompts, and outperforms the baseline qualitatively and quantitatively on long and complicated prompts. To our best knowledge, the proposed method is the first attempt at incorporating large-scale pre-trained diffusion models and distillation sampling for text-driven image generator domain adaptation and gives a quality previously beyond possible. Moreover, we extend our work to 3D-aware style-based generators and DreamBooth guidance.
- NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as General Image Priors | [code]
2D-to-3D reconstruction is an ill-posed problem, yet humans are good at solving this problem due to their prior knowledge of the 3D world developed over years. Driven by this observation, we propose NeRDi, a single-view NeRF synthesis framework with general image priors from 2D diffusion models. Formulating single-view reconstruction as an image-conditioned 3D generation problem, we optimize the NeRF representations by minimizing a diffusion loss on its arbitrary view renderings with a pretrained image diffusion model under the input-view constraint. We leverage off-the-shelf vision-language models and introduce a two-section language guidance as conditioning inputs to the diffusion model. This is essentially helpful for improving multiview content coherence as it narrows down the general image prior conditioned on the semantic and visual features of the single-view input image. Additionally, we introduce a geometric loss based on estimated depth maps to regularize the underlying 3D geometry of the NeRF. Experimental results on the DTU MVS dataset show that our method can synthesize novel views with higher quality even compared to existing methods trained on this dataset. We also demonstrate our generalizability in zero-shot NeRF synthesis for in-the-wild images.
- StegaNeRF: Embedding Invisible Information within Neural Radiance Fields |
[code]Recent advances in neural rendering imply a future of widespread visual data distributions through sharing NeRF model weights. However, while common visual data (images and videos) have standard approaches to embed ownership or copyright information explicitly or subtly, the problem remains unexplored for the emerging NeRF format. We present StegaNeRF, a method for steganographic information embedding in NeRF renderings. We design an optimization framework allowing accurate hidden information extractions from images rendered by NeRF, while preserving its original visual quality. We perform experimental evaluations of our method under several potential deployment scenarios, and we further discuss the insights discovered through our analysis. StegaNeRF signifies an initial exploration into the novel problem of instilling customizable, imperceptible, and recoverable information to NeRF renderings, with minimal impact to rendered images. Project page: this https URL.
- LatentSwap3D: Semantic Edits on 3D Image GANs |
[code]Recent 3D-aware GANs rely on volumetric rendering techniques to disentangle the pose and appearance of objects, de facto generating entire 3D volumes rather than single-view 2D images from a latent code. Complex image editing tasks can be performed in standard 2D-based GANs (e.g., StyleGAN models) as manipulation of latent dimensions. However, to the best of our knowledge, similar properties have only been partially explored for 3D-aware GAN models. This work aims to fill this gap by showing the limitations of existing methods and proposing LatentSwap3D, a model-agnostic approach designed to enable attribute editing in the latent space of pre-trained 3D-aware GANs. We first identify the most relevant dimensions in the latent space of the model controlling the targeted attribute by relying on the feature importance ranking of a random forest classifier. Then, to apply the transformation, we swap the top-K most relevant latent dimensions of the image being edited with an image exhibiting the desired attribute. Despite its simplicity, LatentSwap3D provides remarkable semantic edits in a disentangled manner and outperforms alternative approaches both qualitatively and quantitatively. We demonstrate our semantic edit approach on various 3D-aware generative models such as pi-GAN, GIRAFFE, StyleSDF, MVCGAN, EG3D and VolumeGAN, and on diverse datasets, such as FFHQ, AFHQ, Cats, MetFaces, and CompCars. The project page can be found: \url{this https URL}.
- DiffRF: Rendering-Guided 3D Radiance Field Diffusion | [code]
We introduce DiffRF, a novel approach for 3D radiance field synthesis based on denoising diffusion probabilistic models. While existing diffusion-based methods operate on images, latent codes, or point cloud data, we are the first to directly generate volumetric radiance fields. To this end, we propose a 3D denoising model which directly operates on an explicit voxel grid representation. However, as radiance fields generated from a set of posed images can be ambiguous and contain artifacts, obtaining ground truth radiance field samples is non-trivial. We address this challenge by pairing the denoising formulation with a rendering loss, enabling our model to learn a deviated prior that favours good image quality instead of trying to replicate fitting errors like floating artifacts. In contrast to 2D-diffusion models, our model learns multi-view consistent priors, enabling free-view synthesis and accurate shape generation. Compared to 3D GANs, our diffusion-based approach naturally enables conditional generation such as masked completion or single-view 3D synthesis at inference time.
- SparseFusion: Distilling View-conditioned Diffusion for 3D Reconstruction | [code]
We propose SparseFusion, a sparse view 3D reconstruction approach that unifies recent advances in neural rendering and probabilistic image generation. Existing approaches typically build on neural rendering with re-projected features but fail to generate unseen regions or handle uncertainty under large viewpoint changes. Alternate methods treat this as a (probabilistic) 2D synthesis task, and while they can generate plausible 2D images, they do not infer a consistent underlying 3D. However, we find that this trade-off between 3D consistency and probabilistic image generation does not need to exist. In fact, we show that geometric consistency and generative inference can be complementary in a mode-seeking behavior. By distilling a 3D consistent scene representation from a view-conditioned latent diffusion model, we are able to recover a plausible 3D representation whose renderings are both accurate and realistic. We evaluate our approach across 51 categories in the CO3D dataset and show that it outperforms existing methods, in both distortion and perception metrics, for sparse-view novel view synthesis.
- Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation | [code]
A diffusion model learns to predict a vector field of gradients. We propose to apply chain rule on the learned gradients, and back-propagate the score of a diffusion model through the Jacobian of a differentiable renderer, which we instantiate to be a voxel radiance field. This setup aggregates 2D scores at multiple camera viewpoints into a 3D score, and repurposes a pretrained 2D model for 3D data generation. We identify a technical challenge of distribution mismatch that arises in this application, and propose a novel estimation mechanism to resolve it. We run our algorithm on several off-the-shelf diffusion image generative models, including the recently released Stable Diffusion trained on the large-scale LAION dataset.
- 3D-LDM: Neural Implicit 3D Shape Generation with Latent Diffusion Models | [code]
Diffusion models have shown great promise for image generation, beating GANs in terms of generation diversity, with comparable image quality. However, their application to 3D shapes has been limited to point or voxel representations that can in practice not accurately represent a 3D surface. We propose a diffusion model for neural implicit representations of 3D shapes that operates in the latent space of an auto-decoder. This allows us to generate diverse and high quality 3D surfaces. We additionally show that we can condition our model on images or text to enable image-to-3D generation and text-to-3D generation using CLIP embeddings. Furthermore, adding noise to the latent codes of existing shapes allows us to explore shape variations.
- Efficient Neural Radiance Fields for Interactive Free-viewpoint Video, SIGGRAPH-Asia2022 | [code]
This paper aims to tackle the challenge of efficiently producing interactive free-viewpoint videos. Some recent works equip neural radiance fields with image encoders, enabling them to generalize across scenes. When processing dynamic scenes, they can simply treat each video frame as an individual scene and perform novel view synthesis to generate free-viewpoint videos. However, their rendering process is slow and cannot support interactive applications. A major factor is that they sample lots of points in empty space when inferring radiance fields. We propose a novel scene representation, called ENeRF, for the fast creation of interactive free-viewpoint videos. Specifically, given multi-view images at one frame, we first build the cascade cost volume to predict the coarse geometry of the scene. The coarse geometry allows us to sample few points near the scene surface, thereby significantly improving the rendering speed. This process is fully differentiable, enabling us to jointly learn the depth prediction and radiance field networks from RGB images. Experiments on multiple benchmarks show that our approach exhibits competitive performance while being at least 60 times faster than previous generalizable radiance field methods.
- A Light Touch Approach to Teaching Transformers Multi-view Geometry | [code]
Transformers are powerful visual learners, in large part due to their conspicuous lack of manually-specified priors. This flexibility can be problematic in tasks that involve multiple-view geometry, due to the near-infinite possible variations in 3D shapes and viewpoints (requiring flexibility), and the precise nature of projective geometry (obeying rigid laws). To resolve this conundrum, we propose a "light touch" approach, guiding visual Transformers to learn multiple-view geometry but allowing them to break free when needed. We achieve this by using epipolar lines to guide the Transformer's cross-attention maps, penalizing attention values outside the epipolar lines and encouraging higher attention along these lines since they contain geometrically plausible matches. Unlike previous methods, our proposal does not require any camera pose information at test-time. We focus on pose-invariant object instance retrieval, where standard Transformer networks struggle, due to the large differences in viewpoint between query and retrieved images. Experimentally, our method outperforms state-of-the-art approaches at object retrieval, without needing pose information at test-time.
- High-fidelity 3D GAN Inversion by Pseudo-multi-view Optimization |
[code]We present a high-fidelity 3D generative adversarial network (GAN) inversion framework that can synthesize photo-realistic novel views while preserving specific details of the input image. High-fidelity 3D GAN inversion is inherently challenging due to the geometry-texture trade-off in high-fidelity 3D inversion, where overfitting to a single view input image often damages the estimated geometry during the latent optimization. To solve this challenge, we propose a novel pipeline that builds on the pseudo-multi-view estimation with visibility analysis. We keep the original textures for the visible parts and utilize generative priors for the occluded parts. Extensive experiments show that our approach achieves advantageous reconstruction and novel view synthesis quality over state-of-the-art methods, even for images with out-of-distribution textures. The proposed pipeline also enables image attribute editing with the inverted latent code and 3D-aware texture modification. Our approach enables high-fidelity 3D rendering from a single image, which is promising for various applications of AI-generated 3D content.
- Unsupervised Continual Semantic Adaptation through Neural Rendering | [code]
An increasing amount of applications rely on data-driven models that are deployed for perception tasks across a sequence of scenes. Due to the mismatch between training and deployment data, adapting the model on the new scenes is often crucial to obtain good performance. In this work, we study continual multi-scene adaptation for the task of semantic segmentation, assuming that no ground-truth labels are available during deployment and that performance on the previous scenes should be maintained. We propose training a Semantic-NeRF network for each scene by fusing the predictions of a segmentation model and then using the view-consistent rendered semantic labels as pseudo-labels to adapt the model. Through joint training with the segmentation model, the Semantic-NeRF model effectively enables 2D-3D knowledge transfer. Furthermore, due to its compact size, it can be stored in a long-term memory and subsequently used to render data from arbitrary viewpoints to reduce forgetting. We evaluate our approach on ScanNet, where we outperform both a voxel-based baseline and a state-of-the-art unsupervised domain adaptation method.
- ShadowNeuS: Neural SDF Reconstruction by Shadow Ray Supervision | [code]
By supervising camera rays between a scene and multi-view image planes, NeRF reconstructs a neural scene representation for the task of novel view synthesis. On the other hand, shadow rays between the light source and the scene have yet to be considered. Therefore, we propose a novel shadow ray supervision scheme that optimizes both the samples along the ray and the ray location. By supervising shadow rays, we successfully reconstruct a neural SDF of the scene from single-view pure shadow or RGB images under multiple lighting conditions. Given single-view binary shadows, we train a neural network to reconstruct a complete scene not limited by the camera's line of sight. By further modeling the correlation between the image colors and the shadow rays, our technique can also be effectively extended to RGB inputs. We compare our method with previous works on challenging tasks of shape reconstruction from single-view binary shadow or RGB images and observe significant improvements. The code and data will be released.
- Peekaboo: Text to Image Diffusion Models are Zero-Shot Segmentors | [code]
Recent diffusion-based generative models combined with vision-language models are capable of creating realistic images from natural language prompts. While these models are trained on large internet-scale datasets, such pre-trained models are not directly introduced to any semantic localization or grounding. Most current approaches for localization or grounding rely on human-annotated localization information in the form of bounding boxes or segmentation masks. The exceptions are a few unsupervised methods that utilize architectures or loss functions geared towards localization, but they need to be trained separately. In this work, we explore how off-the-shelf diffusion models, trained with no exposure to such localization information, are capable of grounding various semantic phrases with no segmentation-specific re-training. An inference time optimization process is introduced, that is capable of generating segmentation masks conditioned on natural language. We evaluate our proposal Peekaboo for unsupervised semantic segmentation on the Pascal VOC dataset. In addition, we evaluate for referring segmentation on the RefCOCO dataset. In summary, we present a first zero-shot, open-vocabulary, unsupervised (no localization information), semantic grounding technique leveraging diffusion-based generative models with no re-training. Our code will be released publicly.
- PANeRF: Pseudo-view Augmentation for Improved Neural Radiance Fields Based on Few-shot Inputs | [code]
The method of neural radiance fields (NeRF) has been developed in recent years, and this technology has promising applications for synthesizing novel views of complex scenes. However, NeRF requires dense input views, typically numbering in the hundreds, for generating high-quality images. With a decrease in the number of input views, the rendering quality of NeRF for unseen viewpoints tends to degenerate drastically. To overcome this challenge, we propose pseudo-view augmentation of NeRF, a scheme that expands a sufficient amount of data by considering the geometry of few-shot inputs. We first initialized the NeRF network by leveraging the expanded pseudo-views, which efficiently minimizes uncertainty when rendering unseen views. Subsequently, we fine-tuned the network by utilizing sparse-view inputs containing precise geometry and color information. Through experiments under various settings, we verified that our model faithfully synthesizes novel-view images of superior quality and outperforms existing methods for multi-view datasets.
- Zero NeRF: Registration with Zero Overlap | [code]
We present Zero-NeRF, a projective surface registration method that, to the best of our knowledge, offers the first general solution capable of alignment between scene representations with minimal or zero visual correspondence. To do this, we enforce consistency between visible surfaces of partial and complete reconstructions, which allows us to constrain occluded geometry. We use a NeRF as our surface representation and the NeRF rendering pipeline to perform this alignment. To demonstrate the efficacy of our method, we register real-world scenes from opposite sides with infinitesimal overlaps that cannot be accurately registered using prior methods, and we compare these results against widely used registration methods.
- SPARF: Neural Radiance Fields from Sparse and Noisy Poses | [code]
Neural Radiance Field (NeRF) has recently emerged as a powerful representation to synthesize photorealistic novel views. While showing impressive performance, it relies on the availability of dense input views with highly accurate camera poses, thus limiting its application in real-world scenarios. In this work, we introduce Sparse Pose Adjusting Radiance Field (SPARF), to address the challenge of novel-view synthesis given only few wide-baseline input images (as low as 3) with noisy camera poses. Our approach exploits multi-view geometry constraints in order to jointly learn the NeRF and refine the camera poses. By relying on pixel matches extracted between the input views, our multi-view correspondence objective enforces the optimized scene and camera poses to converge to a global and geometrically accurate solution. Our depth consistency loss further encourages the reconstructed scene to be consistent from any viewpoint. Our approach sets a new state of the art in the sparse-view regime on multiple challenging datasets.
- Magic3D: High-Resolution Text-to-3D Content Creation | [code]
DreamFusion has recently demonstrated the utility of a pre-trained text-to-image diffusion model to optimize Neural Radiance Fields (NeRF), achieving remarkable text-to-3D synthesis results. However, the method has two inherent limitations: (a) extremely slow optimization of NeRF and (b) low-resolution image space supervision on NeRF, leading to low-quality 3D models with a long processing time. In this paper, we address these limitations by utilizing a two-stage optimization framework. First, we obtain a coarse model using a low-resolution diffusion prior and accelerate with a sparse 3D hash grid structure. Using the coarse representation as the initialization, we further optimize a textured 3D mesh model with an efficient differentiable renderer interacting with a high-resolution latent diffusion model. Our method, dubbed Magic3D, can create high quality 3D mesh models in 40 minutes, which is 2x faster than DreamFusion (reportedly taking 1.5 hours on average), while also achieving higher resolution. User studies show 61.7% raters to prefer our approach over DreamFusion. Together with the image-conditioned generation capabilities, we provide users with new ways to control 3D synthesis, opening up new avenues to various creative applications.
- RenderDiffusion: Image Diffusion for 3D Reconstruction, Inpainting and Generation | [code]
Diffusion models currently achieve state-of-the-art performance for both conditional and unconditional image generation. However, so far, image diffusion models do not support tasks required for 3D understanding, such as view-consistent 3D generation or single-view object reconstruction. In this paper, we present RenderDiffusion as the first diffusion model for 3D generation and inference that can be trained using only monocular 2D supervision. At the heart of our method is a novel image denoising architecture that generates and renders an intermediate three-dimensional representation of a scene in each denoising step. This enforces a strong inductive structure into the diffusion process that gives us a 3D consistent representation while only requiring 2D supervision. The resulting 3D representation can be rendered from any viewpoint. We evaluate RenderDiffusion on ShapeNet and Clevr datasets and show competitive performance for generation of 3D scenes and inference of 3D scenes from 2D images. Additionally, our diffusion-based approach allows us to use 2D inpainting to edit 3D scenes. We believe that our work promises to enable full 3D generation at scale when trained on massive image collections, thus circumventing the need to have large-scale 3D model collections for supervision.
- Common Pets in 3D: Dynamic New-View Synthesis of Real-Life Deformable Categories | [code]
Obtaining photorealistic reconstructions of objects from sparse views is inherently ambiguous and can only be achieved by learning suitable reconstruction priors. Earlier works on sparse rigid object reconstruction successfully learned such priors from large datasets such as CO3D. In this paper, we extend this approach to dynamic objects. We use cats and dogs as a representative example and introduce Common Pets in 3D (CoP3D), a collection of crowd-sourced videos showing around 4,200 distinct pets. CoP3D is one of the first large-scale datasets for benchmarking non-rigid 3D reconstruction "in the wild". We also propose Tracker-NeRF, a method for learning 4D reconstruction from our dataset. At test time, given a small number of video frames of an unseen object, Tracker-NeRF predicts the trajectories of its 3D points and generates new views, interpolating viewpoint and time. Results on CoP3D reveal significantly better non-rigid new-view synthesis performance than existing baselines.
- Neural Grasp Distance Fields for Robot Manipulation | [code]
We formulate grasp learning as a neural field and present Neural Grasp Distance Fields (NGDF). Here, the input is a 6D pose of a robot end effector and output is a distance to a continuous manifold of valid grasps for an object. In contrast to current approaches that predict a set of discrete candidate grasps, the distance-based NGDF representation is easily interpreted as a cost, and minimizing this cost produces a successful grasp pose. This grasp distance cost can be incorporated directly into a trajectory optimizer for joint optimization with other costs such as trajectory smoothness and collision avoidance. During optimization, as the various costs are balanced and minimized, the grasp target is allowed to smoothly vary, as the learned grasp field is continuous. In simulation benchmarks with a Franka arm, we find that joint grasping and planning with NGDF outperforms baselines by 63% execution success while generalizing to unseen query poses and unseen object shapes. Project page: this https URL.
- Compressing Explicit Voxel Grid Representations: fast NeRFs become also small | [code]
NeRFs have revolutionized the world of per-scene radiance field reconstruction because of their intrinsic compactness. One of the main limitations of NeRFs is their slow rendering speed, both at training and inference time. Recent research focuses on the optimization of an explicit voxel grid (EVG) that represents the scene, which can be paired with neural networks to learn radiance fields. This approach significantly enhances the speed both at train and inference time, but at the cost of large memory occupation. In this work we propose Re:NeRF, an approach that specifically targets EVG-NeRFs compressibility, aiming to reduce memory storage of NeRF models while maintaining comparable performance. We benchmark our approach with three different EVG-NeRF architectures on four popular benchmarks, showing Re:NeRF's broad usability and effectiveness.
- TANGO: Text-driven Photorealistic and Robust 3D Stylization via Lighting Decomposition, NeurIPS2022 |
[code]Creation of 3D content by stylization is a promising yet challenging problem in computer vision and graphics research. In this work, we focus on stylizing photorealistic appearance renderings of a given surface mesh of arbitrary topology. Motivated by the recent surge of cross-modal supervision of the Contrastive Language-Image Pre-training (CLIP) model, we propose TANGO, which transfers the appearance style of a given 3D shape according to a text prompt in a photorealistic manner. Technically, we propose to disentangle the appearance style as the spatially varying bidirectional reflectance distribution function, the local geometric variation, and the lighting condition, which are jointly optimized, via supervision of the CLIP loss, by a spherical Gaussians based differentiable renderer. As such, TANGO enables photorealistic 3D style transfer by automatically predicting reflectance effects even for bare, low-quality meshes, without training on a task-specific dataset. Extensive experiments show that TANGO outperforms existing methods of text-driven 3D style transfer in terms of photorealistic quality, consistency of 3D geometry, and robustness when stylizing low-quality meshes. Our codes and results are available at our project webpage this https URL.
- Coordinates Are NOT Lonely -- Codebook Prior Helps Implicit Neural 3D Representations, NeurIPS2022 | [code]
Implicit neural 3D representation has achieved impressive results in surface or scene reconstruction and novel view synthesis, which typically uses the coordinate-based multi-layer perceptrons (MLPs) to learn a continuous scene representation. However, existing approaches, such as Neural Radiance Field (NeRF) and its variants, usually require dense input views (i.e. 50-150) to obtain decent results. To relive the over-dependence on massive calibrated images and enrich the coordinate-based feature representation, we explore injecting the prior information into the coordinate-based network and introduce a novel coordinate-based model, CoCo-INR, for implicit neural 3D representation. The cores of our method are two attention modules: codebook attention and coordinate attention. The former extracts the useful prototypes containing rich geometry and appearance information from the prior codebook, and the latter propagates such prior information into each coordinate and enriches its feature representation for a scene or object surface. With the help of the prior information, our method can render 3D views with more photo-realistic appearance and geometries than the current methods using fewer calibrated images available. Experiments on various scene reconstruction datasets, including DTU and BlendedMVS, and the full 3D head reconstruction dataset, H3DS, demonstrate the robustness under fewer input views and fine detail-preserving capability of our proposed method.
- AniFaceGAN: Animatable 3D-Aware Face Image Generation for Video Avatars, NeurIPS2022 |
[code]Although 2D generative models have made great progress in face image generation and animation, they often suffer from undesirable artifacts such as 3D inconsistency when rendering images from different camera viewpoints. This prevents them from synthesizing video animations indistinguishable from real ones. Recently, 3D-aware GANs extend 2D GANs for explicit disentanglement of camera pose by leveraging 3D scene representations. These methods can well preserve the 3D consistency of the generated images across different views, yet they cannot achieve fine-grained control over other attributes, among which facial expression control is arguably the most useful and desirable for face animation. In this paper, we propose an animatable 3D-aware GAN for multiview consistent face animation generation. The key idea is to decompose the 3D representation of the 3D-aware GAN into a template field and a deformation field, where the former represents different identities with a canonical expression, and the latter characterizes expression variations of each identity. To achieve meaningful control over facial expressions via deformation, we propose a 3D-level imitative learning scheme between the generator and a parametric 3D face model during adversarial training of the 3D-aware GAN. This helps our method achieve high-quality animatable face image generation with strong visual 3D consistency, even though trained with only unstructured 2D images. Extensive experiments demonstrate our superior performance over prior works. Project page: this https URL
- LION: Latent Point Diffusion Models for 3D Shape Generation, NeurIPS2022 |
[code]Denoising diffusion models (DDMs) have shown promising results in 3D point cloud synthesis. To advance 3D DDMs and make them useful for digital artists, we require (i) high generation quality, (ii) flexibility for manipulation and applications such as conditional synthesis and shape interpolation, and (iii) the ability to output smooth surfaces or meshes. To this end, we introduce the hierarchical Latent Point Diffusion Model (LION) for 3D shape generation. LION is set up as a variational autoencoder (VAE) with a hierarchical latent space that combines a global shape latent representation with a point-structured latent space. For generation, we train two hierarchical DDMs in these latent spaces. The hierarchical VAE approach boosts performance compared to DDMs that operate on point clouds directly, while the point-structured latents are still ideally suited for DDM-based modeling. Experimentally, LION achieves state-of-the-art generation performance on multiple ShapeNet benchmarks. Furthermore, our VAE framework allows us to easily use LION for different relevant tasks: LION excels at multimodal shape denoising and voxel-conditioned synthesis, and it can be adapted for text- and image-driven 3D generation. We also demonstrate shape autoencoding and latent shape interpolation, and we augment LION with modern surface reconstruction techniques to generate smooth 3D meshes. We hope that LION provides a powerful tool for artists working with 3D shapes due to its high-quality generation, flexibility, and surface reconstruction. Project page and code: this https URL.
- CLIP-Fields: Weakly Supervised Semantic Fields for Robotic Memory | [code]
We propose CLIP-Fields, an implicit scene model that can be trained with no direct human supervision. This model learns a mapping from spatial locations to semantic embedding vectors. The mapping can then be used for a variety of tasks, such as segmentation, instance identification, semantic search over space, and view localization. Most importantly, the mapping can be trained with supervision coming only from web-image and web-text trained models such as CLIP, Detic, and Sentence-BERT. When compared to baselines like Mask-RCNN, our method outperforms on few-shot instance identification or semantic segmentation on the HM3D dataset with only a fraction of the examples. Finally, we show that using CLIP-Fields as a scene memory, robots can perform semantic navigation in real-world environments. Our code and demonstrations are available here: https://mahis.life/clip-fields/
![[code]](https://yuewuhkust.github.io/AniFaceGAN/files/github_icon.jpeg){kind=link}
- Self-improving Multiplane-to-layer Images for Novel View Synthesis, WACV2023 |
[code]We present a new method for lightweight novel-view synthesis that generalizes to an arbitrary forward-facing scene. Recent approaches are computationally expensive, require per-scene optimization, or produce a memory-expensive representation. We start by representing the scene with a set of fronto-parallel semitransparent planes and afterward convert them to deformable layers in an end-to-end manner. Additionally, we employ a feed-forward refinement procedure that corrects the estimated representation by aggregating information from input views. Our method does not require fine-tuning when a new scene is processed and can handle an arbitrary number of views without restrictions. Experimental results show that our approach surpasses recent models in terms of common metrics and human evaluation, with the noticeable advantage in inference speed and compactness of the inferred layered geometry, see this https URL
- Uncertainty-Driven Active Vision for Implicit Scene Reconstruction | [code]
Multi-view implicit scene reconstruction methods have become increasingly popular due to their ability to represent complex scene details. Recent efforts have been devoted to improving the representation of input information and to reducing the number of views required to obtain high quality reconstructions. Yet, perhaps surprisingly, the study of which views to select to maximally improve scene understanding remains largely unexplored. We propose an uncertainty-driven active vision approach for implicit scene reconstruction, which leverages occupancy uncertainty accumulated across the scene using volume rendering to select the next view to acquire. To this end, we develop an occupancy-based reconstruction method which accurately represents scenes using either 2D or 3D supervision. We evaluate our proposed approach on the ABC dataset and the in the wild CO3D dataset, and show that: (1) we are able to obtain high quality state-of-the-art occupancy reconstructions; (2) our perspective conditioned uncertainty definition is effective to drive improvements in next best view selection and outperforms strong baseline approaches; and (3) we can further improve shape understanding by performing a gradient-based search on the view selection candidates. Overall, our results highlight the importance of view selection for implicit scene reconstruction, making it a promising avenue to explore further.
- SinGRAV: Learning a Generative Radiance Volume from a Single Natural Scene | [code]
We present a 3D generative model for general natural scenes. Lacking necessary volumes of 3D data characterizing the target scene, we propose to learn from a single scene. Our key insight is that a natural scene often contains multiple constituents whose geometry, texture, and spatial arrangements follow some clear patterns, but still exhibit rich variations over different regions within the same scene. This suggests localizing the learning of a generative model on substantial local regions. Hence, we exploit a multi-scale convolutional network, which possesses the spatial locality bias in nature, to learn from the statistics of local regions at multiple scales within a single scene. In contrast to existing methods, our learning setup bypasses the need to collect data from many homogeneous 3D scenes for learning common features. We coin our method SinGRAV, for learning a Generative RAdiance Volume from a Single natural scene. We demonstrate the ability of SinGRAV in generating plausible and diverse variations from a single scene, the merits of SinGRAV over state-of-the-art generative neural scene methods, as well as the versatility of SinGRAV by its use in a variety of applications, spanning 3D scene editing, composition, and animation. Code and data will be released to facilitate further research.
- IntrinsicNeRF: Learning Intrinsic Neural Radiance Fields for Editable Novel View Synthesis |
[code]We present intrinsic neural radiance fields, dubbed IntrinsicNeRF, that introduce intrinsic decomposition into the NeRF-based~\cite{mildenhall2020nerf} neural rendering method and can perform editable novel view synthesis in room-scale scenes while existing inverse rendering combined with neural rendering methods~\cite{zhang2021physg, zhang2022modeling} can only work on object-specific scenes. Given that intrinsic decomposition is a fundamentally ambiguous and under-constrained inverse problem, we propose a novel distance-aware point sampling and adaptive reflectance iterative clustering optimization method that enables IntrinsicNeRF with traditional intrinsic decomposition constraints to be trained in an unsupervised manner, resulting in temporally consistent intrinsic decomposition results. To cope with the problem of different adjacent instances of similar reflectance in a scene being incorrectly clustered together, we further propose a hierarchical clustering method with coarse-to-fine optimization to obtain a fast hierarchical indexing representation. It enables compelling real-time augmented reality applications such as scene recoloring, material editing, and illumination variation. Extensive experiments on Blender Object and Replica Scene demonstrate that we can obtain high-quality, consistent intrinsic decomposition results and high-fidelity novel view synthesis even for challenging sequences. Code and data are available on the project webpage: this https URL.
- Structure-Aware NeRF without Posed Camera via Epipolar Constraint |
[code]The neural radiance field (NeRF) for realistic novel view synthesis requires camera poses to be pre-acquired by a structure-from-motion (SfM) approach. This two-stage strategy is not convenient to use and degrades the performance because the error in the pose extraction can propagate to the view synthesis. We integrate the pose extraction and view synthesis into a single end-to-end procedure so they can benefit from each other. For training NeRF models, only RGB images are given, without pre-known camera poses. The camera poses are obtained by the epipolar constraint in which the identical feature in different views has the same world coordinates transformed from the local camera coordinates according to the extracted poses. The epipolar constraint is jointly optimized with pixel color constraint. The poses are represented by a CNN-based deep network, whose input is the related frames. This joint optimization enables NeRF to be aware of the scene's structure that has an improved generalization performance. Extensive experiments on a variety of scenes demonstrate the effectiveness of the proposed approach. Code is available at this https URL.
- Improving 3D-aware Image Synthesis with A Geometry-aware Discriminator, NeurIPS2022 |
[code]3D-aware image synthesis aims at learning a generative model that can render photo-realistic 2D images while capturing decent underlying 3D shapes. A popular solution is to adopt the generative adversarial network (GAN) and replace the generator with a 3D renderer, where volume rendering with neural radiance field (NeRF) is commonly used. Despite the advancement of synthesis quality, existing methods fail to obtain moderate 3D shapes. We argue that, considering the two-player game in the formulation of GANs, only making the generator 3D-aware is not enough. In other words, displacing the generative mechanism only offers the capability, but not the guarantee, of producing 3D-aware images, because the supervision of the generator primarily comes from the discriminator. To address this issue, we propose GeoD through learning a geometry-aware discriminator to improve 3D-aware GANs. Concretely, besides differentiating real and fake samples from the 2D image space, the discriminator is additionally asked to derive the geometry information from the inputs, which is then applied as the guidance of the generator. Such a simple yet effective design facilitates learning substantially more accurate 3D shapes. Extensive experiments on various generator architectures and training datasets verify the superiority of GeoD over state-of-the-art alternatives. Moreover, our approach is registered as a general framework such that a more capable discriminator (i.e., with a third task of novel view synthesis beyond domain classification and geometry extraction) can further assist the generator with a better multi-view consistency.
- MonoNeuralFusion: Online Monocular Neural 3D Reconstruction with Geometric Priors | [code]
High-fidelity 3D scene reconstruction from monocular videos continues to be challenging, especially for complete and fine-grained geometry reconstruction. The previous 3D reconstruction approaches with neural implicit representations have shown a promising ability for complete scene reconstruction, while their results are often over-smooth and lack enough geometric details. This paper introduces a novel neural implicit scene representation with volume rendering for high-fidelity online 3D scene reconstruction from monocular videos. For fine-grained reconstruction, our key insight is to incorporate geometric priors into both the neural implicit scene representation and neural volume rendering, thus leading to an effective geometry learning mechanism based on volume rendering optimization. Benefiting from this, we present MonoNeuralFusion to perform the online neural 3D reconstruction from monocular videos, by which the 3D scene geometry is efficiently generated and optimized during the on-the-fly 3D monocular scanning. The extensive comparisons with state-of-the-art approaches show that our MonoNeuralFusion consistently generates much better complete and fine-grained reconstruction results, both quantitatively and qualitatively.
- SymmNeRF: Learning to Explore Symmetry Prior for Single-View View Synthesis, ACCV2022 |
[code]We study the problem of novel view synthesis of objects from a single image. Existing methods have demonstrated the potential in single-view view synthesis. However, they still fail to recover the fine appearance details, especially in self-occluded areas. This is because a single view only provides limited information. We observe that manmade objects usually exhibit symmetric appearances, which introduce additional prior knowledge. Motivated by this, we investigate the potential performance gains of explicitly embedding symmetry into the scene representation. In this paper, we propose SymmNeRF, a neural radiance field (NeRF) based framework that combines local and global conditioning under the introduction of symmetry priors. In particular, SymmNeRF takes the pixel-aligned image features and the corresponding symmetric features as extra inputs to the NeRF, whose parameters are generated by a hypernetwork. As the parameters are conditioned on the image-encoded latent codes, SymmNeRF is thus scene-independent and can generalize to new scenes. Experiments on synthetic and realworld datasets show that SymmNeRF synthesizes novel views with more details regardless of the pose transformation, and demonstrates good generalization when applied to unseen objects. Code is available at: this https URL.
- 360FusionNeRF: Panoramic Neural Radiance Fields with Joint Guidance | [code]
We present a method to synthesize novel views from a single 360∘ panorama image based on the neural radiance field (NeRF). Prior studies in a similar setting rely on the neighborhood interpolation capability of multi-layer perceptions to complete missing regions caused by occlusion, which leads to artifacts in their predictions. We propose 360FusionNeRF, a semi-supervised learning framework where we introduce geometric supervision and semantic consistency to guide the progressive training process. Firstly, the input image is re-projected to 360∘ images, and auxiliary depth maps are extracted at other camera positions. The depth supervision, in addition to the NeRF color guidance, improves the geometry of the synthesized views. Additionally, we introduce a semantic consistency loss that encourages realistic renderings of novel views. We extract these semantic features using a pre-trained visual encoder such as CLIP, a Vision Transformer trained on hundreds of millions of diverse 2D photographs mined from the web with natural language supervision. Experiments indicate that our proposed method can produce plausible completions of unobserved regions while preserving the features of the scene. When trained across various scenes, 360FusionNeRF consistently achieves the state-of-the-art performance when transferring to synthetic Structured3D dataset (PSNR
5%, SSIM3% LPIPS13%), real-world Matterport3D dataset (PSNR3%, SSIM3% LPIPS9%) and Replica360 dataset (PSNR8%, SSIM2% LPIPS~18%).
- PNeRF: Probabilistic Neural Scene Representations for Uncertain 3D Visual Mapping, ICRA2023 | [code]
Recently neural scene representations have provided very impressive results for representing 3D scenes visually, however, their study and progress have mainly been limited to visualization of virtual models in computer graphics or scene reconstruction in computer vision without explicitly accounting for sensor and pose uncertainty. Using this novel scene representation in robotics applications, however, would require accounting for this uncertainty in the neural map. The aim of this paper is therefore to propose a novel method for training {\em probabilistic neural scene representations} with uncertain training data that could enable the inclusion of these representations in robotics applications. Acquiring images using cameras or depth sensors contains inherent uncertainty, and furthermore, the camera poses used for learning a 3D model are also imperfect. If these measurements are used for training without accounting for their uncertainty, then the resulting models are non-optimal, and the resulting scene representations are likely to contain artifacts such as blur and un-even geometry. In this work, the problem of uncertainty integration to the learning process is investigated by focusing on training with uncertain information in a probabilistic manner. The proposed method involves explicitly augmenting the training likelihood with an uncertainty term such that the learnt probability distribution of the network is minimized with respect to the training uncertainty. It will be shown that this leads to more accurate image rendering quality, in addition to more precise and consistent geometry. Validation has been carried out on both synthetic and real datasets showing that the proposed approach outperforms state-of-the-art methods. The results show notably that the proposed method is capable of rendering novel high-quality views even when the training data is limited.
- ActiveNeRF: Learning where to See with Uncertainty Estimation |
[code]Recently, Neural Radiance Fields (NeRF) has shown promising performances on reconstructing 3D scenes and synthesizing novel views from a sparse set of 2D images. Albeit effective, the performance of NeRF is highly influenced by the quality of training samples. With limited posed images from the scene, NeRF fails to generalize well to novel views and may collapse to trivial solutions in unobserved regions. This makes NeRF impractical under resource-constrained scenarios. In this paper, we present a novel learning framework, ActiveNeRF, aiming to model a 3D scene with a constrained input budget. Specifically, we first incorporate uncertainty estimation into a NeRF model, which ensures robustness under few observations and provides an interpretation of how NeRF understands the scene. On this basis, we propose to supplement the existing training set with newly captured samples based on an active learning scheme. By evaluating the reduction of uncertainty given new inputs, we select the samples that bring the most information gain. In this way, the quality of novel view synthesis can be improved with minimal additional resources. Extensive experiments validate the performance of our model on both realistic and synthetic scenes, especially with scarcer training data. Code will be released at \url{this https URL}.
- Learning A Unified 3D Point Cloud for View Synthesis | [code]
3D point cloud representation-based view synthesis methods have demonstrated effectiveness. However, existing methods usually synthesize novel views only from a single source view, and it is non-trivial to generalize them to handle multiple source views for pursuing higher reconstruction quality. In this paper, we propose a new deep learning-based view synthesis paradigm, which learns a unified 3D point cloud from different source views. Specifically, we first construct sub-point clouds by projecting source views to 3D space based on their depth maps. Then, we learn the unified 3D point cloud by adaptively fusing points at a local neighborhood defined on the union of the sub-point clouds. Besides, we also propose a 3D geometry-guided image restoration module to fill the holes and recover high-frequency details of the rendered novel views. Experimental results on three benchmark datasets demonstrate that our method outperforms state-of-the-art view synthesis methods to a large extent both quantitatively and visually.
- Dual-Space NeRF: Learning Animatable Avatars and Scene Lighting in Separate Spaces, 3DV2022 | [code]
Modeling the human body in a canonical space is a common practice for capturing and animation. But when involving the neural radiance field (NeRF), learning a static NeRF in the canonical space is not enough because the lighting of the body changes when the person moves even though the scene lighting is constant. Previous methods alleviate the inconsistency of lighting by learning a per-frame embedding, but this operation does not generalize to unseen poses. Given that the lighting condition is static in the world space while the human body is consistent in the canonical space, we propose a dual-space NeRF that models the scene lighting and the human body with two MLPs in two separate spaces. To bridge these two spaces, previous methods mostly rely on the linear blend skinning (LBS) algorithm. However, the blending weights for LBS of a dynamic neural field are intractable and thus are usually memorized with another MLP, which does not generalize to novel poses. Although it is possible to borrow the blending weights of a parametric mesh such as SMPL, the interpolation operation introduces more artifacts. In this paper, we propose to use the barycentric mapping, which can directly generalize to unseen poses and surprisingly achieves superior results than LBS with neural blending weights. Quantitative and qualitative results on the Human3.6M and the ZJU-MoCap datasets show the effectiveness of our method.
- DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation | [code]
Large text-to-image models achieved a remarkable leap in the evolution of AI, enabling high-quality and diverse synthesis of images from a given text prompt. However, these models lack the ability to mimic the appearance of subjects in a given reference set and synthesize novel renditions of them in different contexts. In this work, we present a new approach for "personalization" of text-to-image diffusion models (specializing them to users' needs). Given as input just a few images of a subject, we fine-tune a pretrained text-to-image model (Imagen, although our method is not limited to a specific model) such that it learns to bind a unique identifier with that specific subject. Once the subject is embedded in the output domain of the model, the unique identifier can then be used to synthesize fully-novel photorealistic images of the subject contextualized in different scenes. By leveraging the semantic prior embedded in the model with a new autogenous class-specific prior preservation loss, our technique enables synthesizing the subject in diverse scenes, poses, views, and lighting conditions that do not appear in the reference images. We apply our technique to several previously-unassailable tasks, including subject recontextualization, text-guided view synthesis, appearance modification, and artistic rendering (all while preserving the subject's key features). Project page: this https URL
- E-NeRF: Neural Radiance Fields from a Moving Event Camera | [code]
Estimating neural radiance fields (NeRFs) from ideal images has been extensively studied in the computer vision community. Most approaches assume optimal illumination and slow camera motion. These assumptions are often violated in robotic applications, where images contain motion blur and the scene may not have suitable illumination. This can cause significant problems for downstream tasks such as navigation, inspection or visualization of the scene. To alleviate these problems we present E-NeRF, the first method which estimates a volumetric scene representation in the form of a NeRF from a fast-moving event camera. Our method can recover NeRFs during very fast motion and in high dynamic range conditions, where frame-based approaches fail. We show that rendering high-quality frames is possible by only providing an event stream as input. Furthermore, by combining events and frames, we can estimate NeRFs of higher quality than state-of-the-art approaches under severe motion blur. We also show that combining events and frames can overcome failure cases of NeRF estimation in scenarios where only few input views are available, without requiring additional regularization.
- FurryGAN: High Quality Foreground-aware Image Synthesis, ECCV2022 |
[code]Foreground-aware image synthesis aims to generate images as well as their foreground masks. A common approach is to formulate an image as an masked blending of a foreground image and a background image. It is a challenging problem because it is prone to reach the trivial solution where either image overwhelms the other, i.e., the masks become completely full or empty, and the foreground and background are not meaningfully separated. We present FurryGAN with three key components: 1) imposing both the foreground image and the composite image to be realistic, 2) designing a mask as a combination of coarse and fine masks, and 3) guiding the generator by an auxiliary mask predictor in the discriminator. Our method produces realistic images with remarkably detailed alpha masks which cover hair, fur, and whiskers in a fully unsupervised manner.
- Text-to-Image Generation via Implicit Visual Guidance and Hypernetwork | [code]
We develop an approach for text-to-image generation that embraces additional retrieval images, driven by a combination of implicit visual guidance loss and generative objectives. Unlike most existing text-to-image generation methods which merely take the text as input, our method dynamically feeds cross-modal search results into a unified training stage, hence improving the quality, controllability and diversity of generation results. We propose a novel hypernetwork modulated visual-text encoding scheme to predict the weight update of the encoding layer, enabling effective transfer from visual information (e.g. layout, content) into the corresponding latent domain. Experimental results show that our model guided with additional retrieval visual data outperforms existing GAN-based models. On COCO dataset, we achieve better FID of 9.13 with up to 3.5× fewer generator parameters, compared with the state-of-the-art method.
- UPST-NeRF: Universal Photorealistic Style Transfer of Neural Radiance Fields for 3D Scene |
[code]3D scenes photorealistic stylization aims to generate photorealistic images from arbitrary novel views according to a given style image while ensuring consistency when rendering from different viewpoints. Some existing stylization methods with neural radiance fields can effectively predict stylized scenes by combining the features of the style image with multi-view images to train 3D scenes. However, these methods generate novel view images that contain objectionable artifacts. Besides, they cannot achieve universal photorealistic stylization for a 3D scene. Therefore, a styling image must retrain a 3D scene representation network based on a neural radiation field. We propose a novel 3D scene photorealistic style transfer framework to address these issues. It can realize photorealistic 3D scene style transfer with a 2D style image. We first pre-trained a 2D photorealistic style transfer network, which can meet the photorealistic style transfer between any given content image and style image. Then, we use voxel features to optimize a 3D scene and get the geometric representation of the scene. Finally, we jointly optimize a hyper network to realize the scene photorealistic style transfer of arbitrary style images. In the transfer stage, we use a pre-trained 2D photorealistic network to constrain the photorealistic style of different views and different style images in the 3D scene. The experimental results show that our method not only realizes the 3D photorealistic style transfer of arbitrary style images but also outperforms the existing methods in terms of visual quality and consistency. Project page:this https URL.
- HRF-Net: Holistic Radiance Fields from Sparse Inputs | [code]
We present HRF-Net, a novel view synthesis method based on holistic radiance fields that renders novel views using a set of sparse inputs. Recent generalizing view synthesis methods also leverage the radiance fields but the rendering speed is not real-time. There are existing methods that can train and render novel views efficiently but they can not generalize to unseen scenes. Our approach addresses the problem of real-time rendering for generalizing view synthesis and consists of two main stages: a holistic radiance fields predictor and a convolutional-based neural renderer. This architecture infers not only consistent scene geometry based on the implicit neural fields but also renders new views efficiently using a single GPU. We first train HRF-Net on multiple 3D scenes of the DTU dataset and the network can produce plausible novel views on unseen real and synthetics data using only photometric losses. Moreover, our method can leverage a denser set of reference images of a single scene to produce accurate novel views without relying on additional explicit representations and still maintains the high-speed rendering of the pre-trained model. Experimental results show that HRF-Net outperforms state-of-the-art generalizable neural rendering methods on various synthetic and real datasets.
- NeSF: Neural Semantic Fields for Generalizable Semantic Segmentation of 3D Scenes | [code]
We present NeSF, a method for producing 3D semantic fields from pre-trained density fields and sparse 2D semantic supervision. Our method side-steps traditional scene representations by leveraging neural representations where 3D information is stored within neural fields. In spite of being supervised by 2D signals alone, our method is able to generate 3D-consistent semantic maps from novel camera poses and can be queried at arbitrary 3D points. Notably, NeSF is compatible with any method producing a density field, and its accuracy improves as the quality of the pre-trained density fields improve. Our empirical analysis demonstrates comparable quality to competitive 2D and 3D semantic segmentation baselines on convincing synthetic scenes while also offering features unavailable to existing methods.
- Transformers as Meta-Learners for Implicit Neural Representations, ECCV2022 |
[code]Implicit Neural Representations (INRs) have emerged and shown their benefits over discrete representations in recent years. However, fitting an INR to the given observations usually requires optimization with gradient descent from scratch, which is inefficient and does not generalize well with sparse observations. To address this problem, most of the prior works train a hypernetwork that generates a single vector to modulate the INR weights, where the single vector becomes an information bottleneck that limits the reconstruction precision of the output INR. Recent work shows that the whole set of weights in INR can be precisely inferred without the single-vector bottleneck by gradient-based meta-learning. Motivated by a generalized formulation of gradient-based meta-learning, we propose a formulation that uses Transformers as hypernetworks for INRs, where it can directly build the whole set of INR weights with Transformers specialized as set-to-set mapping. We demonstrate the effectiveness of our method for building INRs in different tasks and domains, including 2D image regression and view synthesis for 3D objects. Our work draws connections between the Transformer hypernetworks and gradient-based meta-learning algorithms and we provide further analysis for understanding the generated INRs.
- VolTeMorph: Realtime, Controllable and Generalisable Animation of Volumetric Representations | [code]
The recent increase in popularity of volumetric representations for scene reconstruction and novel view synthesis has put renewed focus on animating volumetric content at high visual quality and in real-time. While implicit deformation methods based on learned functions can produce impressive results, they are `black boxes' to artists and content creators, they require large amounts of training data to generalise meaningfully, and they do not produce realistic extrapolations outside the training data. In this work we solve these issues by introducing a volume deformation method which is real-time, easy to edit with off-the-shelf software and can extrapolate convincingly. To demonstrate the versatility of our method, we apply it in two scenarios: physics-based object deformation and telepresence where avatars are controlled using blendshapes. We also perform thorough experiments showing that our method compares favourably to both volumetric approaches combined with implicit deformation and methods based on mesh deformation.
- ZEPI-Net: Light Field Super Resolution via Internal Cross-Scale Epipolar Plane Image Zero-Shot Learning, Neural Processing Letters (2022) | [code]
Many applications of light field (LF) imaging have been limited by the spatial-angular resolution problem, hence the need for efficient super-resolution techniques. Recently, learning-based solutions have achieved remarkably better performances than traditional super-resolution (SR) techniques. Unfortunately, the learning or training process relies heavily on the training dataset, which could be limited for most LF imaging applications. In this paper, we propose a novel LF spatial-angular SR algorithm based on zero-shot learning. We suggest learning cross-scale reusable features in the epipolar plane image (EPI) space, and avoiding explicitly modeling scene priors or implicitly learning that from a large number of LFs. Most importantly, without using any external LFs, the proposed algorithm can simultaneously super-resolve a LF in both spatial and angular domains. Moreover, the proposed solution is free of depth or disparity estimation, which is usually employed by existing LF spatial and angular SR. By using a simple 8-layers fully convolutional network, we show that the proposed algorithm can generate comparable results to the state-of-the-art spatial SR. Our algorithm outperforms the existing methods in terms of angular SR on multiple groups of public LF datasets. The experiment results indicate that the cross-scale features can be well learned and be reused for LF SR in the EPI space.
- ObjectFusion: Accurate object-level SLAM with neural object priors, Graphical Models, Volume 123, September 2022 | [code]
Previous object-level Simultaneous Localization and Mapping (SLAM) approaches still fail to create high quality object-oriented 3D map in an efficient way. The main challenges come from how to represent the object shape effectively and how to apply such object representation to accurate online camera tracking efficiently. In this paper, we provide ObjectFusion as a novel object-level SLAM in static scenes which efficiently creates object-oriented 3D map with high-quality object reconstruction, by leveraging neural object priors. We propose a neural object representation with only a single encoder–decoder network to effectively express the object shape across various categories, which benefits high quality reconstruction of object instance. More importantly, we propose to convert such neural object representation as precise measurements to jointly optimize the object shape, object pose and camera pose for the final accurate 3D object reconstruction. With extensive evaluations on synthetic and real-world RGB-D datasets, we show that our ObjectFusion outperforms previous approaches, with better object reconstruction quality, using much less memory footprint, and in a more efficient way, especially at the object level.
- End-to-end View Synthesis via NeRF Attention | [code]
In this paper, we present a simple seq2seq formulation for view synthesis where we take a set of ray points as input and output colors corresponding to the rays. Directly applying a standard transformer on this seq2seq formulation has two limitations. First, the standard attention cannot successfully fit the volumetric rendering procedure, and therefore high-frequency components are missing in the synthesized views. Second, applying global attention to all rays and pixels is extremely inefficient. Inspired by the neural radiance field (NeRF), we propose the NeRF attention (NeRFA) to address the above problems. On the one hand, NeRFA considers the volumetric rendering equation as a soft feature modulation procedure. In this way, the feature modulation enhances the transformers with the NeRF-like inductive bias. On the other hand, NeRFA performs multi-stage attention to reduce the computational overhead. Furthermore, the NeRFA model adopts the ray and pixel transformers to learn the interactions between rays and pixels. NeRFA demonstrates superior performance over NeRF and NerFormer on four datasets: DeepVoxels, Blender, LLFF, and CO3D. Besides, NeRFA establishes a new state-of-the-art under two settings: the single-scene view synthesis and the category-centric novel view synthesis. The code will be made publicly available.
- CLA-NeRF: Category-Level Articulated Neural Radiance Field, ICRA2022 | [code]
We propose CLA-NeRF -- a Category-Level Articulated Neural Radiance Field that can perform view synthesis, part segmentation, and articulated pose estimation. CLA-NeRF is trained at the object category level using no CAD models and no depth, but a set of RGB images with ground truth camera poses and part segments. During inference, it only takes a few RGB views (i.e., few-shot) of an unseen 3D object instance within the known category to infer the object part segmentation and the neural radiance field. Given an articulated pose as input, CLA-NeRF can perform articulation-aware volume rendering to generate the corresponding RGB image at any camera pose. Moreover, the articulated pose of an object can be estimated via inverse rendering. In our experiments, we evaluate the framework across five categories on both synthetic and real-world data. In all cases, our method shows realistic deformation results and accurate articulated pose estimation. We believe that both few-shot articulated object rendering and articulated pose estimation open doors for robots to perceive and interact with unseen articulated objects.
- GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis, NeurIPS2020 |
[code]While 2D generative adversarial networks have enabled high-resolution image synthesis, they largely lack an understanding of the 3D world and the image formation process. Thus, they do not provide precise control over camera viewpoint or object pose. To address this problem, several recent approaches leverage intermediate voxel-based representations in combination with differentiable rendering. However, existing methods either produce low image resolution or fall short in disentangling camera and scene properties, eg, the object identity may vary with the viewpoint. In this paper, we propose a generative model for radiance fields which have recently proven successful for novel view synthesis of a single scene. In contrast to voxel-based representations, radiance fields are not confined to a coarse discretization of the 3D space, yet allow for disentangling camera and scene properties while degrading gracefully in the presence of reconstruction ambiguity. By introducing a multi-scale patch-based discriminator, we demonstrate synthesis of high-resolution images while training our model from unposed 2D images alone. We systematically analyze our approach on several challenging synthetic and real-world datasets. Our experiments reveal that radiance fields are a powerful representation for generative image synthesis, leading to 3D consistent models that render with high fidelity.
- GRF: Learning a General Radiance Field for 3D Scene Representation and Rendering, ICCV2021(oral) |
[code]We present a simple yet powerful neural network that implicitly represents and renders 3D objects and scenes only from 2D observations. The network models 3D geometries as a general radiance field, which takes a set of 2D images with camera poses and intrinsics as input, constructs an internal representation for each point of the 3D space, and then renders the corresponding appearance and geometry of that point viewed from an arbitrary position. The key to our approach is to learn local features for each pixel in 2D images and to then project these features to 3D points, thus yielding general and rich point representations. We additionally integrate an attention mechanism to aggregate pixel features from multiple 2D views, such that visual occlusions are implicitly taken into account. Extensive experiments demonstrate that our method can generate high-quality and realistic novel views for novel objects, unseen categories and challenging real-world scenes.
- pixelNeRF: Neural Radiance Fields from One or Few Images, CVPR2021 |
[code]We propose pixelNeRF, a learning framework that predicts a continuous neural scene representation conditioned on one or few input images. The existing approach for constructing neural radiance fields involves optimizing the representation to every scene independently, requiring many calibrated views and significant compute time. We take a step towards resolving these shortcomings by introducing an architecture that conditions a NeRF on image inputs in a fully convolutional manner. This allows the network to be trained across multiple scenes to learn a scene prior, enabling it to perform novel view synthesis in a feed-forward manner from a sparse set of views (as few as one). Leveraging the volume rendering approach of NeRF, our model can be trained directly from images with no explicit 3D supervision. We conduct extensive experiments on ShapeNet benchmarks for single image novel view synthesis tasks with held-out objects as well as entire unseen categories. We further demonstrate the flexibility of pixelNeRF by demonstrating it on multi-object ShapeNet scenes and real scenes from the DTU dataset. In all cases, pixelNeRF outperforms current state-of-the-art baselines for novel view synthesis and single image 3D reconstruction. For the video and code, please visit the project website: this https URL
- Learned Initializations for Optimizing Coordinate-Based Neural Representations, CVPR2021 |
[code]Coordinate-based neural representations have shown significant promise as an alternative to discrete, array-based representations for complex low dimensional signals. However, optimizing a coordinate-based network from randomly initialized weights for each new signal is inefficient. We propose applying standard meta-learning algorithms to learn the initial weight parameters for these fully-connected networks based on the underlying class of signals being represented (e.g., images of faces or 3D models of chairs). Despite requiring only a minor change in implementation, using these learned initial weights enables faster convergence during optimization and can serve as a strong prior over the signal class being modeled, resulting in better generalization when only partial observations of a given signal are available. We explore these benefits across a variety of tasks, including representing 2D images, reconstructing CT scans, and recovering 3D shapes and scenes from 2D image observations.
- pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesis, CVPR2021(oral) |
[code]We have witnessed rapid progress on 3D-aware image synthesis, leveraging recent advances in generative visual models and neural rendering. Existing approaches however fall short in two ways: first, they may lack an underlying 3D representation or rely on view-inconsistent rendering, hence synthesizing images that are not multi-view consistent; second, they often depend upon representation network architectures that are not expressive enough, and their results thus lack in image quality. We propose a novel generative model, named Periodic Implicit Generative Adversarial Networks (π-GAN or pi-GAN), for high-quality 3D-aware image synthesis. π-GAN leverages neural representations with periodic activation functions and volumetric rendering to represent scenes as view-consistent 3D representations with fine detail. The proposed approach obtains state-of-the-art results for 3D-aware image synthesis with multiple real and synthetic datasets.
- Portrait Neural Radiance Fields from a Single Image | [code]
We present a method for estimating Neural Radiance Fields (NeRF) from a single headshot portrait. While NeRF has demonstrated high-quality view synthesis, it requires multiple images of static scenes and thus impractical for casual captures and moving subjects. In this work, we propose to pretrain the weights of a multilayer perceptron (MLP), which implicitly models the volumetric density and colors, with a meta-learning framework using a light stage portrait dataset. To improve the generalization to unseen faces, we train the MLP in the canonical coordinate space approximated by 3D face morphable models. We quantitatively evaluate the method using controlled captures and demonstrate the generalization to real portrait images, showing favorable results against state-of-the-arts.
- CAMPARI: Camera-Aware Decomposed Generative Neural Radiance Fields | [code]
Tremendous progress in deep generative models has led to photorealistic image synthesis. While achieving compelling results, most approaches operate in the two-dimensional image domain, ignoring the three-dimensional nature of our world. Several recent works therefore propose generative models which are 3D-aware, i.e., scenes are modeled in 3D and then rendered differentiably to the image plane. This leads to impressive 3D consistency, but incorporating such a bias comes at a price: the camera needs to be modeled as well. Current approaches assume fixed intrinsics and a predefined prior over camera pose ranges. As a result, parameter tuning is typically required for real-world data, and results degrade if the data distribution is not matched. Our key hypothesis is that learning a camera generator jointly with the image generator leads to a more principled approach to 3D-aware image synthesis. Further, we propose to decompose the scene into a background and foreground model, leading to more efficient and disentangled scene representations. While training from raw, unposed image collections, we learn a 3D- and camera-aware generative model which faithfully recovers not only the image but also the camera data distribution. At test time, our model generates images with explicit control over the camera as well as the shape and appearance of the scene.
- NeRF-VAE: A Geometry Aware 3D Scene Generative Model | [code]
We propose NeRF-VAE, a 3D scene generative model that incorporates geometric structure via NeRF and differentiable volume rendering. In contrast to NeRF, our model takes into account shared structure across scenes, and is able to infer the structure of a novel scene -- without the need to re-train -- using amortized inference. NeRF-VAE's explicit 3D rendering process further contrasts previous generative models with convolution-based rendering which lacks geometric structure. Our model is a VAE that learns a distribution over radiance fields by conditioning them on a latent scene representation. We show that, once trained, NeRF-VAE is able to infer and render geometrically-consistent scenes from previously unseen 3D environments using very few input images. We further demonstrate that NeRF-VAE generalizes well to out-of-distribution cameras, while convolutional models do not. Finally, we introduce and study an attention-based conditioning mechanism of NeRF-VAE's decoder, which improves model performance.
- Unconstrained Scene Generation with Locally Conditioned Radiance Fields, ICCV2021 |
[code]We follow an adversarial learning framework, where the generator models scenes via their radiance field, and the discriminator attempts to distinguish between images rendered from those radiance fields and images of real scenes. Conceptually, our model decomposes the radiance field of a scene into many small local radiance fields that result from conditioning on a 2D grid of latent codes W. W can be interpreted as a latent floorplan representing the scene.
- MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereo, ICCV2021 |
[code]We present MVSNeRF, a novel neural rendering approach that can efficiently reconstruct neural radiance fields for view synthesis. Unlike prior works on neural radiance fields that consider per-scene optimization on densely captured images, we propose a generic deep neural network that can reconstruct radiance fields from only three nearby input views via fast network inference. Our approach leverages plane-swept cost volumes (widely used in multi-view stereo) for geometry-aware scene reasoning, and combines this with physically based volume rendering for neural radiance field reconstruction. We train our network on real objects in the DTU dataset, and test it on three different datasets to evaluate its effectiveness and generalizability. Our approach can generalize across scenes (even indoor scenes, completely different from our training scenes of objects) and generate realistic view synthesis results using only three input images, significantly outperforming concurrent works on generalizable radiance field reconstruction. Moreover, if dense images are captured, our estimated radiance field representation can be easily fine-tuned; this leads to fast per-scene reconstruction with higher rendering quality and substantially less optimization time than NeRF.
- Stereo Radiance Fields (SRF): Learning View Synthesis from Sparse Views of Novel Scenes, CVPR2021 |
[code]Recent neural view synthesis methods have achieved impressive quality and realism, surpassing classical pipelines which rely on multi-view reconstruction. State-of-the-Art methods, such as NeRF, are designed to learn a single scene with a neural network and require dense multi-view inputs. Testing on a new scene requires re-training from scratch, which takes 2-3 days. In this work, we introduce Stereo Radiance Fields (SRF), a neural view synthesis approach that is trained end-to-end, generalizes to new scenes, and requires only sparse views at test time. The core idea is a neural architecture inspired by classical multi-view stereo methods, which estimates surface points by finding similar image regions in stereo images. In SRF, we predict color and density for each 3D point given an encoding of its stereo correspondence in the input images. The encoding is implicitly learned by an ensemble of pair-wise similarities -- emulating classical stereo. Experiments show that SRF learns structure instead of overfitting on a scene. We train on multiple scenes of the DTU dataset and generalize to new ones without re-training, requiring only 10 sparse and spread-out views as input. We show that 10-15 minutes of fine-tuning further improve the results, achieving significantly sharper, more detailed results than scene-specific models. The code, model, and videos are available at this https URL.
- Neural Rays for Occlusion-aware Image-based Rendering, CVPR2022 |
[code]We present a new neural representation, called Neural Ray (NeuRay), for the novel view synthesis task. Recent works construct radiance fields from image features of input views to render novel view images, which enables the generalization to new scenes. However, due to occlusions, a 3D point may be invisible to some input views. On such a 3D point, these generalization methods will include inconsistent image features from invisible views, which interfere with the radiance field construction. To solve this problem, we predict the visibility of 3D points to input views within our NeuRay representation. This visibility enables the radiance field construction to focus on visible image features, which significantly improves its rendering quality. Meanwhile, a novel consistency loss is proposed to refine the visibility in NeuRay when finetuning on a specific scene. Experiments demonstrate that our approach achieves state-of-the-art performance on the novel view synthesis task when generalizing to unseen scenes and outperforms per-scene optimization methods after finetuning.
- Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis, ICCV2021 |
[code]We present DietNeRF, a 3D neural scene representation estimated from a few images. Neural Radiance Fields (NeRF) learn a continuous volumetric representation of a scene through multi-view consistency, and can be rendered from novel viewpoints by ray casting. While NeRF has an impressive ability to reconstruct geometry and fine details given many images, up to 100 for challenging 360° scenes, it often finds a degenerate solution to its image reconstruction objective when only a few input views are available. To improve few-shot quality, we propose DietNeRF. We introduce an auxiliary semantic consistency loss that encourages realistic renderings at novel poses. DietNeRF is trained on individual scenes to (1) correctly render given input views from the same pose, and (2) match high-level semantic attributes across different, random poses. Our semantic loss allows us to supervise DietNeRF from arbitrary poses. We extract these semantics using a pre-trained visual encoder such as CLIP, a Vision Transformer trained on hundreds of millions of diverse single-view, 2D photographs mined from the web with natural language supervision. In experiments, DietNeRF improves the perceptual quality of few-shot view synthesis when learned from scratch, can render novel views with as few as one observed image when pre-trained on a multi-view dataset, and produces plausible completions of completely unobserved regions.
- CodeNeRF: Disentangled Neural Radiance Fields for Object Categories, ICCV2021(oral) |
[code]CodeNeRF is an implicit 3D neural representation that learns the variation of object shapes and textures across a category and can be trained, from a set of posed images, to synthesize novel views of unseen objects. Unlike the original NeRF, which is scene specific, CodeNeRF learns to disentangle shape and texture by learning separate embeddings. At test time, given a single unposed image of an unseen object, CodeNeRF jointly estimates camera viewpoint, and shape and appearance codes via optimization. Unseen objects can be reconstructed from a single image, and then rendered from new viewpoints or their shape and texture edited by varying the latent codes. We conduct experiments on the SRN benchmark, which show that CodeNeRF generalises well to unseen objects and achieves on-par performance with methods that require known camera pose at test time. Our results on real-world images demonstrate that CodeNeRF can bridge the sim-to-real gap.
- StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image Synthesis, ICLR2022 |
[code]We propose StyleNeRF, a 3D-aware generative model for photo-realistic high-resolution image synthesis with high multi-view consistency, which can be trained on unstructured 2D images. Existing approaches either cannot synthesize high-resolution images with fine details or yield noticeable 3D-inconsistent artifacts. In addition, many of them lack control over style attributes and explicit 3D camera poses. StyleNeRF integrates the neural radiance field (NeRF) into a style-based generator to tackle the aforementioned challenges, i.e., improving rendering efficiency and 3D consistency for high-resolution image generation. We perform volume rendering only to produce a low-resolution feature map and progressively apply upsampling in 2D to address the first issue. To mitigate the inconsistencies caused by 2D upsampling, we propose multiple designs, including a better upsampler and a new regularization loss. With these designs, StyleNeRF can synthesize high-resolution images at interactive rates while preserving 3D consistency at high quality. StyleNeRF also enables control of camera poses and different levels of styles, which can generalize to unseen views. It also supports challenging tasks, including zoom-in and-out, style mixing, inversion, and semantic editing.
- GNeRF: GAN-based Neural Radiance Field without Posed Camera, ICCV2021(oral) | [code]
We introduce GNeRF, a framework to marry Generative Adversarial Networks (GAN) with Neural Radiance Field (NeRF) reconstruction for the complex scenarios with unknown and even randomly initialized camera poses. Recent NeRF-based advances have gained popularity for remarkable realistic novel view synthesis. However, most of them heavily rely on accurate camera poses estimation, while few recent methods can only optimize the unknown camera poses in roughly forward-facing scenes with relatively short camera trajectories and require rough camera poses initialization. Differently, our GNeRF only utilizes randomly initialized poses for complex outside-in scenarios. We propose a novel two-phases end-to-end framework. The first phase takes the use of GANs into the new realm for optimizing coarse camera poses and radiance fields jointly, while the second phase refines them with additional photometric loss. We overcome local minima using a hybrid and iterative optimization scheme. Extensive experiments on a variety of synthetic and natural scenes demonstrate the effectiveness of GNeRF. More impressively, our approach outperforms the baselines favorably in those scenes with repeated patterns or even low textures that are regarded as extremely challenging before.
- NeRD: Neural Reflectance Decomposition from Image Collections, ICCV2021 |
[code]Decomposing a scene into its shape, reflectance, and illumination is a challenging but important problem in computer vision and graphics. This problem is inherently more challenging when the illumination is not a single light source under laboratory conditions but is instead an unconstrained environmental illumination. Though recent work has shown that implicit representations can be used to model the radiance field of an object, most of these techniques only enable view synthesis and not relighting. Additionally, evaluating these radiance fields is resource and time-intensive. We propose a neural reflectance decomposition (NeRD) technique that uses physically-based rendering to decompose the scene into spatially varying BRDF material properties. In contrast to existing techniques, our input images can be captured under different illumination conditions. In addition, we also propose techniques to convert the learned reflectance volume into a relightable textured mesh enabling fast real-time rendering with novel illuminations. We demonstrate the potential of the proposed approach with experiments on both synthetic and real datasets, where we are able to obtain high-quality relightable 3D assets from image collections.
- NeRF++: Analyzing and Improving Neural Radiance Fields |
[code]Neural Radiance Fields (NeRF) achieve impressive view synthesis results for a variety of capture settings, including 360 capture of bounded scenes and forward-facing capture of bounded and unbounded scenes. NeRF fits multi-layer perceptrons (MLPs) representing view-invariant opacity and view-dependent color volumes to a set of training images, and samples novel views based on volume rendering techniques. In this technical report, we first remark on radiance fields and their potential ambiguities, namely the shape-radiance ambiguity, and analyze NeRF's success in avoiding such ambiguities. Second, we address a parametrization issue involved in applying NeRF to 360 captures of objects within large-scale, unbounded 3D scenes. Our method improves view synthesis fidelity in this challenging scenario. Code is available at this https URL.
- GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields, CVPR2021(oral) |
[code]Deep generative models allow for photorealistic image synthesis at high resolutions. But for many applications, this is not enough: content creation also needs to be controllable. While several recent works investigate how to disentangle underlying factors of variation in the data, most of them operate in 2D and hence ignore that our world is three-dimensional. Further, only few works consider the compositional nature of scenes. Our key hypothesis is that incorporating a compositional 3D scene representation into the generative model leads to more controllable image synthesis. Representing scenes as compositional generative neural feature fields allows us to disentangle one or multiple objects from the background as well as individual objects' shapes and appearances while learning from unstructured and unposed image collections without any additional supervision. Combining this scene representation with a neural rendering pipeline yields a fast and realistic image synthesis model. As evidenced by our experiments, our model is able to disentangle individual objects and allows for translating and rotating them in the scene as well as changing the camera pose.
- FiG-NeRF: Figure Ground Neural Radiance Fields for 3D Object Category Modelling, 3DV2021 | [code]
We investigate the use of Neural Radiance Fields (NeRF) to learn high quality 3D object category models from collections of input images. In contrast to previous work, we are able to do this whilst simultaneously separating foreground objects from their varying backgrounds. We achieve this via a 2-component NeRF model, FiG-NeRF, that prefers explanation of the scene as a geometrically constant background and a deformable foreground that represents the object category. We show that this method can learn accurate 3D object category models using only photometric supervision and casually captured images of the objects. Additionally, our 2-part decomposition allows the model to perform accurate and crisp amodal segmentation. We quantitatively evaluate our method with view synthesis and image fidelity metrics, using synthetic, lab-captured, and in-the-wild data. Our results demonstrate convincing 3D object category modelling that exceed the performance of existing methods.
- NerfingMVS: Guided Optimization of Neural Radiance Fields for Indoor Multi-view Stereo, ICCV2021(oral) |
[code]In this work, we present a new multi-view depth estimation method that utilizes both conventional SfM reconstruction and learning-based priors over the recently proposed neural radiance fields (NeRF). Unlike existing neural network based optimization method that relies on estimated correspondences, our method directly optimizes over implicit volumes, eliminating the challenging step of matching pixels in indoor scenes. The key to our approach is to utilize the learning-based priors to guide the optimization process of NeRF. Our system firstly adapts a monocular depth network over the target scene by finetuning on its sparse SfM reconstruction. Then, we show that the shape-radiance ambiguity of NeRF still exists in indoor environments and propose to address the issue by employing the adapted depth priors to monitor the sampling process of volume rendering. Finally, a per-pixel confidence map acquired by error computation on the rendered image can be used to further improve the depth quality. Experiments show that our proposed framework significantly outperforms state-of-the-art methods on indoor scenes, with surprising findings presented on the effectiveness of correspondence-based optimization and NeRF-based optimization over the adapted depth priors. In addition, we show that the guided optimization scheme does not sacrifice the original synthesis capability of neural radiance fields, improving the rendering quality on both seen and novel views.