Rate Distortion Characteristic Modeling for Neural Image Compression
This work addresses the practical deployment issue of neural image compression by eliminating the need for multiple models for different bit-rates, though it is incremental in improving existing methods.
The paper tackles the problem of rate-distortion characteristic modeling for neural image compression, enabling a single trained network to achieve arbitrary bit-rate points with state-of-the-art continuous bit-rate coding performance.
End-to-end optimized neural image compression (NIC) has obtained superior lossy compression performance recently. In this paper, we consider the problem of rate-distortion (R-D) characteristic analysis and modeling for NIC. We make efforts to formulate the essential mathematical functions to describe the R-D behavior of NIC using deep networks. Thus arbitrary bit-rate points could be elegantly realized by leveraging such model via a single trained network. We propose a plugin-in module to learn the relationship between the target bit-rate and the binary representation for the latent variable of auto-encoder. The proposed scheme resolves the problem of training distinct models to reach different points in the R-D space. Furthermore, we model the rate and distortion characteristic of NIC as a function of the coding parameter $λ$ respectively. Our experiments show our proposed method is easy to adopt and realizes state-of-the-art continuous bit-rate coding performance, which implies that our approach would benefit the practical deployment of NIC.