# Channel-Level Variable Quantization Network for Deep Image Compression

arXiv: https://arxiv.org/abs/2007.12619 · HTML: https://mlsanity.com/paper/2007.12619/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Jul 2020.

## Built on by (most cited first, 6 of 6)

- [Variable-Rate Deep Image Compression through Spatially-Adaptive Feature Transform](https://mlsanity.com/paper/2108.09551/) (2021)
- [A Survey on Approximate Edge AI for Energy Efficient Autonomous Driving Services](https://mlsanity.com/paper/2304.14271/) (2023)
- [Exploring Structural Sparsity in Neural Image Compression](https://mlsanity.com/paper/2202.04595/) (2022)
- [Comprehensive Comparisons of Uniform Quantization in Deep Image Compression](https://mlsanity.com/paper/2303.00587/) (2023)
- [Rethinking Learned Image Compression: Context is All You Need](https://mlsanity.com/paper/2407.11590/) (2024)
- [HiDE: Hierarchical Dictionary-Based Entropy Modeling for Learned Image Compression](https://mlsanity.com/paper/2603.06766/) (2026)

## Abstract (arXiv)

Deep image compression systems mainly contain four components: encoder, quantizer, entropy model, and decoder. To optimize these four components, a joint rate-distortion framework was proposed, and many deep neural network-based methods achieved great success in image compression. However, almost all convolutional neural network-based methods treat channel-wise feature maps equally, reducing the flexibility in handling different types of information. In this paper, we propose a channel-level variable quantization network to dynamically allocate more bitrates for significant channels and withdraw bitrates for negligible channels. Specifically, we propose a variable quantization controller. It consists of two key components: the channel importance module, which can dynamically learn the importance of channels during training, and the splitting-merging module, which can allocate different bitrates for different channels. We also formulate the quantizer into a Gaussian mixture model manner. Quantitative and qualitative experiments verify the effectiveness of the proposed model and demonstrate that our method achieves superior performance and can produce much better visual reconstructions.