From f87a3e2558cbfd7af2e2ee424ceb12249b1f35a5 Mon Sep 17 00:00:00 2001 From: Jan Blunk Date: Mon, 28 Aug 2023 17:14:19 +0200 Subject: [PATCH] absolute image path --- README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index ff30c75..49738c7 100644 --- a/README.md +++ b/README.md @@ -1,10 +1,10 @@ # Beyond Debiasing: Actively Steering Feature Selection via Loss Regularization ## Overview -This repository provides code to use the method presented in our GCPR 2023 paper **"Beyond Debiasing: Actively Steering Feature Selection via Loss Regularization"**. If you want to get started, take a look at our [example network](https://git.inf-cv.uni-jena.de/blunk/beyond-debiasing/src/main/regression_network.py) and the corresponding [jupyter notebook](https://git.inf-cv.uni-jena.de/blunk/beyond-debiasing/src/main/feature_steering_example.ipynb). +This repository provides code to use the method presented in our GCPR 2023 paper **"Beyond Debiasing: Actively Steering Feature Selection via Loss Regularization"**. If you want to get started, take a look at our [example network](regression_network.py) and the corresponding [jupyter notebook](feature_steering_example.ipynb).
-By measuring the feature usage, we can steer the model towards (not) using features that are specifically (un-)desired. + By measuring the feature usage, we can steer the model towards (not) using features that are specifically (un-)desired.
Our method generalizes from debiasing to the **encouragement and discouragement of arbitrary features**. That is, it not only aims at removing the influence of undesired features / biases but also at increasing the influence of features that are known to be well-established from domain knowledge. @@ -28,7 +28,7 @@ If you use our method, please cite: First, create an environment with pip and Python first (Anaconda environment / Python virtual environment). We recommend to install [PyTorch with CUDA support](https://pytorch.org/get-started/locally/). Then, you can install all subsequent packages via pip as described above. ## Usage in Python -Since our method relies on loss regularization, it is very simple to add to your own networks - you only need to modify your loss function. To help with that, we provide an [exemplary network](https://git.inf-cv.uni-jena.de/blunk/beyond-debiasing/src/main/regression_network.py) and a [jupyter notebook](https://git.inf-cv.uni-jena.de/blunk/beyond-debiasing/src/main/feature_steering_example.ipynb) with example code. +Since our method relies on loss regularization, it is very simple to add to your own networks - you only need to modify your loss function. To help with that, we provide an [exemplary network](regression_network.py) and a [jupyter notebook](feature_steering_example.ipynb) with example code. ## Repository Organization * Installation: