Q-LIC: Quantizing Learned Image Compression with Channel Splitting

TL;DR

This paper introduces QLIC, a channel splitting-based quantization method for learned image compression, reducing network complexity and improving coding gain, making LIC more feasible for resource-limited devices.

Contribution

The paper proposes a novel channel splitting approach to reduce quantization error impact, enhancing LIC performance and hardware suitability.

Findings

BD-rate reduction of up to 4.74% with 8-bit quantization

Better coding gain than state-of-the-art quantization methods

Feasibility of hardware implementation demonstrated

Abstract

Learned image compression (LIC) has reached a comparable coding gain with traditional hand-crafted methods such as VVC intra. However, the large network complexity prohibits the usage of LIC on resource-limited embedded systems. Network quantization is an efficient way to reduce the network burden. This paper presents a quantized LIC (QLIC) by channel splitting. First, we explore that the influence of quantization error to the reconstruction error is different for various channels. Second, we split the channels whose quantization has larger influence to the reconstruction error. After the splitting, the dynamic range of channels is reduced so that the quantization error can be reduced. Finally, we prune several channels to keep the number of overall channels as origin. By using the proposal, in the case of 8-bit quantization for weight and activation of both main and hyper path, we can reduce the BD-rate by 0.61%-4.74% compared with the previous QLIC. Besides, we can reach better coding gain compared with the state-of-the-art network quantization method when quantizing MS-SSIM models. Moreover, our proposal can be combined with other network quantization methods to further improve the coding gain. The moderate coding loss caused by the quantization validates the feasibility of the hardware implementation for QLIC in the future.