Denoising Vision Transformers

[**Jiawei Yang**](https://jiawei-yang.github.io/)^1*† · [**Katie Z Luo**](https://www.cs.cornell.edu/~katieluo/)^2* · [**Jiefeng Li**](https://jeffli.site/)³ · [**Congyue Deng**](https://cs.stanford.edu/~congyue/)⁴
[**Leonidas Guibas**](https://geometry.stanford.edu/member/guibas/)⁴ · [**Dilip Krishnan**](https://dilipkay.wordpress.com/)⁵ · [**Kilian Q. Weinberger**](https://www.cs.cornell.edu/~kilian/)²
[**Yonglong Tian**](https://people.csail.mit.edu/yonglong/)⁵ · [**Yue Wang**](https://yuewang.xyz/)^{1 1}University of Southern California   ²Cornell University
³Shanghai Jiaotong University   ⁴Stanford University
⁵Google Research
†project lead *equal technical contribution contribution Accepted to ECCV 2024

TL;DR

This work presents Denoising Vision Transformers (DVT). It removes the visually annoying artifacts commonly seen in ViTs' feature maps and substaintially improves the downstream performance of dense recognition tasks.

Citation

@article{yang2024denoising,
  author = {Yang, Jiawei and Luo, Katie Z and Li, Jiefeng and Deng, Congyue and Guibas, Leonidas J. and Krishnan, Dilip and Weinberger, Kilian Q and Tian, Yonglong and Wang, Yue},
  title = {DVT: Denoising Vision Transformers},
  journal = {arXiv preprint arXiv:2401.02957},
  year = {2024},
}

This README file and codebase are legacy. We will update them soon.

Installation

1. Create a conda environment.

conda create -n dvt python=3.10 -y

2. Activate the environment.

conda activate dvt

3. Install dependencies from requirements.txt.

pip install -r requirements.txt

4. Install tiny-cuda-nn manually:

pip install git+https://github.com/NVlabs/tiny-cuda-nn/#subdirectory=bindings/torch

If you encounter the error nvcc fatal : Unsupported gpu architecture compute_89, try the following command:

TCNN_CUDA_ARCHITECTURES=86 pip install git+https://github.com/NVlabs/tiny-cuda-nn/#subdirectory=bindings/torch

If you encounter the error: parameter packs not expanded with ‘...’, Refer to this solution on GitHub.

Data preparation

1. PASCAL-VOC 2007 and 2012: Please download the PASCAL VOC07 and PASCAL VOC12 datasets (link) and put the data in the folder data, e.g.,

mkdir -p data
cd data
wget http://host.robots.ox.ac.uk/pascal/VOC/voc2007/VOCtrainval_06-Nov-2007.tar
tar -xf VOCtrainval_06-Nov-2007.tar
wget http://host.robots.ox.ac.uk/pascal/VOC/voc2012/VOCtrainval_11-May-2012.tar
tar -xf VOCtrainval_11-May-2012.tar

In our experiments reported in the paper, we used the first 10,000 examples from data/voc_train.txt for stage-1 denoising. This text file was generated by gathering all JPG images from data/VOC2007/JPEGImages and data/VOC2012/JPEGImages, excluding the validation images, and then randomly shuffling them.

2. ADE20K: [legacy, need to check] Please download the ADE20K dataset and put the data in data/ADEChallengeData2016.

3. NYU-D: Please download the NYU-depth dataset and put the data in data/nyu. Results are provided given the 2014 annotations following previous works.

4. ImageNet (Optional):

   • Download the ImageNet dataset from http://www.image-net.org/
   • Extract data following the instructions at here.
   • Put the data in data/imagenet.

Run the code

See sample_scripts for examples of running the code.

We provide some demo outputs in demo/demo_outputs. For example, this image shows our denoising results on a cat image: From left to right, we show: (1) input crop, (2) raw DINOv2 base output, (3) Kmeans clustering of the raw output, (4) L2 feature norm of the raw output, (5) the similarity between the central patch and other patches in the raw output, (6) our denoised output, (7) Kmeans clustering of the denoised output, (8) L2 feature norm of the denoised output, (9) the similarity between the central patch and other patches in the denoised output, (10) the decomposed shared artifacts, (11) the L2 norm of the shared artifacts, (12) the ground-truth residual error, (13) the predicted residual term, and (13) the composition of the shared artifacts and the predicted residual term.

Main Results and Checkpoints

VOC Evaluation Results

ADE20K Evaluation Results

NYU-D Evaluation Results

Denoiser Checkpoints

[ ] To be released.