KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#dsh #dsh-plugin #typescript #ai-agent #developer-tools #claude-code #codex #deepseek-harness #ai-agents #react #mcp #rust

NIC-RobustBench

★ 50 Open GitHub ↗

A Comprehensive Open-Source Toolkit for Neural Image Compression and Robustness Analysis

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# <img align="left" width="100" height="100" src="imgs/logo.png"> NIC-RobustBench

A Comprehensive Open-Source Toolkit for Neural Image Compression and Robustness Analysis. Please refer to the paper for more details:

![Benchmark scheme](https://raw.githubusercontent.com/msu-video-group/NIC-RobustBench/main/imgs/scheme1.png)

## :sparkles: Features

![Benchmark Features](https://raw.githubusercontent.com/msu-video-group/NIC-RobustBench/main/imgs/scheme2.png)

## :whale: Prerequisites
- Docker 23+ with NVIDIA Container Toolkit
- 40 Gb of GPU RAM
- We provide Dockerfiles with complete list of requirements

![Attack example](https://raw.githubusercontent.com/msu-video-group/NIC-RobustBench/main/imgs/attack_example.png)
<center>
Attack example on JPEG AI v5.1.
</center>

## :zap: Quick Start
Visit [GitLab repository](https://vg-code.gml-team.ru/framework/codecs-robustness) to find out more about GitLab CI/CD launch.

```bash
# Download source code
git pull https://github.com/msu-video-group/NIC-RobustBench
cd NIC-RobustBench

# build docker images
# if you change image names, please change DOCKER_IMAGE and DOCKER_IMAGE_JPEGAI variables accordingly in launch.sh script
sudo docker build -f main.Dockerfile -t codecs_main . 
sudo docker build -f jpegai.Dockerfile -t codecs_jpegai .
# note that building can take up to 20-30 minutes
 
# run attack
chmod +x launch.sh
./launch.sh {attack_preset} {loss_name} {attack_name} {codec_name} {gpu_id}

# example of launching
./launch.sh 0 bpp_increase_loss random-noise jpegai-v51-hop-b05 0

# to launch a set of attacks/codecs you can edit launch_all.sh script
# configure launch_all.sh to run desired attacks/codecs 
chmod +x launch.sh launch_all.sh
./launch_all.sh
```

Be aware that the first time script is launching it will download all necessary weights of about 10 Gb. To download them manually run in the root of directory:
```bash
wget --backups=1 -nv https://titan.gml-team.ru:5003/fsdownload/o3kmmUJdU/models.zip \
         https://titan.gml-team.ru:5003/fsdownload/o3kmmUJdU/models.zip && rm models.zip.1
```

#### Datasets
We include one image from each of BSDS, NIPS2017, Kodak PhotoCD and CityScapes datasets in this repository. You can download these datasets completely via:
```bash
wget --backups=1 -nv https://titan.gml-team.ru:5003/fsdownload/Hy8LfJM6e/codec-datasets.zip https://titan.gml-team.ru:5003/fsdownload/Hy8LfJM6e/codec-datasets.zip && codec-datasets.zip.1 
```

You can add any new dataset in the *datasets/codec-datasets* folder. If you do so, change TEST_DATASET_NAMES and TEST_DATASET_PATHS variables in scripts-docker/env_vars.sh accordingly. **Make sure each dataset subfolder contains at least one image.**

## :zap: Tutorial video
[![NIC-RobustBench video](https://img.youtube.com/vi/SgZ1WezyG2c/0.jpg)](https://www.youtube.com/watch?v=SgZ1WezyG2c)

NIC-RobustBench

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README