1. Info Title: On Disentangled Training for Nonlinear Transform in Learned Image Compression Author: Han Li Pub: ICLR 2025 Repo: https://github.com/qingshi9974/AuxT Keyword: Discrete Wavelet Transform

1. Info Title: Linear Attention Modeling for Learned Image Compression Author: Donghui Feng Pub: CVPR 2025 Repo: https://github.com/sjtu-medialab/RwkvCompress Keyword: RWKV, Checkboard 2. Motivation

Info Title: Learned Image Compression with Mixed Transformer-CNN Architectures Author: Jinming Liu Pub: CVPR 2023 Repo: https://github.com/jmliu206/LIC_TCM Keyword: Transformer, CNN Motivation 局部信息和

一些VSCode中的常见配置 VSCode 设置 Copilot 本地运行 (解决Copilot无响应问题) VSCode Settings –> Extension Kind –> Edit in settings.json –> 添加如下代码: 1234"remote.extensionKind": { "GitH

本次简单讲一下Cheng 2020，也是经典的baseline之一，作者也开源了，推荐阅读。 Info Title: Learned Image Compression with Discretized Gaussian Mixture Likelihoods and Attention Modules Author: Zhengxue Cheng Pub: CVPR 2020 Repo: h

Info Title: Learned Image Compression with Dictionary-based Entropy Model Author: Jingbo Lu Pub: CVPR 2025 Repo: https://github.com/LabShuHangGU/DCAE Keyword: Dictionary-based, Cross Attention Motiva

Info Title: Variational Image Compression with a Scale Hyperprior Author: Johannes Balle Pub: ICLR 2018 Keyword: Hyperprior Motivation 已有研究（例如Ballel2017的工作）还是用的变换+量化的形式来压缩。然而这些方法在量化时往往忽略了图像中的空间依赖性，导

Info Title: End-to-end optimized image compression Author: Johannes Balle Pub: ICLR 2017 Repo: https://github.com/tensorflow/compression Motivation 联合优化率（rate）和失真（distortion）是困难的，现有算法例如JPEG通常是线性变换、均