Learning True Rate-Distortion-Optimization for End-To-End Image Compression
This work addresses the problem of inefficient rate-distortion optimization in end-to-end image compression for applications requiring high compression efficiency, representing an incremental improvement over prior methods.
The paper tackles the challenge of integrating rate-distortion optimization into end-to-end image compression by enhancing training with low-complexity estimations and proposing fast inference modes, achieving average rate savings of 19.6% in MS-SSIM over their previous model and 27.3% over a conventional deep coder.
Even though rate-distortion optimization is a crucial part of traditional image and video compression, not many approaches exist which transfer this concept to end-to-end-trained image compression. Most frameworks contain static compression and decompression models which are fixed after training, so efficient rate-distortion optimization is not possible. In a previous work, we proposed RDONet, which enables an RDO approach comparable to adaptive block partitioning in HEVC. In this paper, we enhance the training by introducing low-complexity estimations of the RDO result into the training. Additionally, we propose fast and very fast RDO inference modes. With our novel training method, we achieve average rate savings of 19.6% in MS-SSIM over the previous RDONet model, which equals rate savings of 27.3% over a comparable conventional deep image coder.