KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#developer-tools #typescript #codex #ai-agents #claude-code #dsh #dsh-plugin #cli #macos #ai-agent #deepseek-harness #open-source

CVQN

★ 30 Open GitHub ↗

The code for Channel-Level Variable Quantization Network for Deep Image Compression (IJCAI 2020)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# CVQN

This repository is for [Channel-Level Variable Quantization Network for Deep Image Compression](https://arxiv.org/abs/2007.12619).

(to appear in IJCAI, 2020)

By [Zhisheng Zhong](https://zzs1994.github.io), Hiroaki Akutsu and [Kiyoharu Aizawa](https://www.hal.t.u-tokyo.ac.jp/~aizawa/).

## Table of contents
- [Overview](#overview)
- [Data Download](#data-download)
- [Folder Structure](#folder-structure)
- [Training and Evaluation](#training-and-evaluation)

***********************************************************************************************************

## Overview

<div align=center><img src="https://github.com/zzs1994/CVQN/blob/master/page_image/overview_CVQN.jpg" width="90%" height="90%"></div align=center>
<p align="center">Framework of the channel-level variable quantization network.</p>

## Dependencies

- Python (3.7.5)
- PyTorch (1.2.0)
- torchvision (0.4.0)
- PyYaml (5.2)
- tensorboard (2.0.1)

## Data Download

These training datasets can be downloaded from the above links.

- [DIV2K](https://data.vision.ee.ethz.ch/cvl/DIV2K)
- [Flickr2K](http://cv.snu.ac.kr/research/EDSR/Flickr2K.tar)
- [CLIC2019](https://www.compression.cc/challenge)

## Folder Structure

Your CVQN folder may be similar to this:

```
-- logs (log folder)
-- ckps (checkpoint folder)
-- tbs (tensorboard log folder)
-- yaml (yaml folder)
-- pytorch_msssim
-- config
-- *.py
```

## Training and Evaluation

Please modify the training & evaluation dataset path in `yaml/XXX.yaml`.

You can also modify other parameters to change the model and training strategy in the same file.

An example to train a model:

```bash
python main_train_eval.py --config yaml/XXX.yaml
```

### Citation

If you find this code useful, please cite our paper:

```
@inproceedings{Zhong2020CVQN,
  title     = {Channel-Level Variable Quantization Network for Deep Image Compression},
  author    = {Zhong, Zhisheng and Akutsu, Hiroaki and Aizawa, Kiyoharu},
  booktitle = {Proceedings of the International Joint Conference on Artificial Intelligence},
  year      = {2020}
}
```

***********************************************************************************************************

CVQN

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README