Image Inpainting application which works on reconstructing a distorted or hazy image to its original form, is becoming a prominent part of research. The existing approaches often suffer from blurry & distorted images due to discontinuity of local pixels. The local pixel discontinuity happens due to ignoring the semantic relevance and feature continuity of hole/missing regions.
In this paper The human behavior in repairing pictures,i.e. two steps as conception and painting are introduced to ensure global structure consistency as well as local pixel continuity of a picture .The approach of painting process,constructing new structures and texture from the end nodes of the structures created previously,that ensures the local pixel continuity of the final result. Inspired by this ,a coherent semantic attention layer (CSA)is proposed which will retain fine details.
-
The Proposed model consists of two steps: 1. Rough Inpainting & 2. Refinement Inpainting
-
1. Rough Inpainting - Training of a rough network to rough out the missing contents/holes is followed. This stabilizes the training and ensures ta larger receptive field.
be the ground truth &
(3* 256 * 256 shape) be the input to rough framework. From these two we try to have the rough prediction
by using
convolutions with skip connections.
-
2. Refinement Inpainting - This consists of two stages
-
Refinement Network -
convolutions to double channels &
-
Coherent Semantic Attention -A novel coherent semantic attention layer is applied to construct the correlation between the deep features of hole regions. Completed in two phases: Searching and Generating
Searching & Generating.
-
Searching - CSA layer searches the closest-matching contextual patch
in known region
to initialize
in feature map are considered &
.Which stands for similarity between
&
.
-
Generating - As a initial patch the left of
initially is 0. as
Based on this,an attention map
is used to create the attention matrix. Which with deconvolutional layers reconstruct
-
-
-
Consistency Loss - A ImageNet-pretrained VGG-16 is used to extract a high level feature space from the original image. In
distance as :
-
Feature Path Discriminator - A feature patch discriminator is developed to discriminate completed images and original images by inspecting their feature maps. As receptive fields of each point features the entire image, the adversarial loss
,
& Loss function
for discriminators:
where
, represents the operation of taking average for all real/fake data in the mini-batch.
- On three datasets this model was tested Places2 , CelebA,and Paris StreetView model is optimized by the Adam algorithm
with a learning rate of 2 × 10−4 and
= 0.5. The tradeoff parameters are set as
= 1,
= 0.01,
= 0.002,with a batch size of 1.The weak correlation areas in attention maps are areas of concern for generated patches which are far from it, the strong correlation areas are areas of concern for both adjacent generated patches and most relevant patches.
- Effectof CSA Layer - The CSA layer was replacced with convolutional layers as a query to understand it's importance . And it failed to perform like CSA as the pixels were not consistent with the background. Hence global semantic structure and local coherency are constructed by the CSA layer.
The inpainting task thus is completed through this two stage method. Whether the unknown region is irregular or centering, this algorithm can achieve state-of-the-art inpainting results. The Human inspired Rough & Refinement steps added new path on this application.
The stated method clearly had great effect on inpainting techniques and worked well than other Architectures it compared with. The new introduced Feature path Discriminator & Consistency loss achieved greater impact than conventional ways.
For code, visit this link.