Skip to content

lin-toto/lic-multistage-spatial-context

Folders and files

NameName
Last commit message
Last commit date

Latest commit

44d65a2 · Jul 17, 2023
Feb 13, 2023
Nov 13, 2020
Jul 17, 2023
May 11, 2021
Apr 15, 2023
Mar 7, 2023
Feb 9, 2023
Nov 23, 2022
Feb 9, 2023
Nov 23, 2022
Dec 11, 2021
Nov 24, 2022
Sep 2, 2020
Dec 11, 2021
Jan 26, 2023
Nov 23, 2022
Sep 1, 2020
Dec 24, 2021
Feb 12, 2021
Jan 20, 2023
Jan 26, 2023
Jul 17, 2023
May 4, 2021
Feb 9, 2023
Feb 10, 2023
Jul 8, 2023

Repository files navigation

Multistage Spatial Context Models for Learned Image Compression

This is the official code repository for the paper Multistage Spatial Context Models for Learned Image Compression.

This implementation is based on CompressAI. Please check the commit history for detailed change logs.

This repository contains:

  • CompressAI commit hash b10cc7c
  • An efficient Gaussian Mixture Model implementation
  • Multistage Spatial Context Model implementations on Cheng2020

Pretrained weights for MSE are available here.

Usage

Installation

Clone this repository and install with:

python3 setup.py install

You may have to remove your existing CompressAI installation first as this implementation replaces it. We recommend creating a new virtual environment for testing this implementation.

Evaluation

Use the CompressAI scripts for evaluating the models:

python3 -m compressai.utils.eval_model pretrained /path/to/dataset -a cheng2020-attn-gmm-multistage-{2x2,4x4} -m mse -q {1-6}

About

Multistage Spatial Context Models for Learned Image Compression

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages