MultiQuan RDP: Rate-Distortion-Perception Coding via Offset Quantizers

Ruida Zhou, Chao Tian

Keywords: information theory, quantization, rate-distortion-perception, compression

TL;DR: We propose the MultiQuan quantizers interpolating between single quantizer and dithered quantization for rate-distortion-perception coding.

Abstract: The rate-distortion-perception (RDP) framework has attracted significant recent attention due to its application in neural compression. It is important to understand the underlying mechanism connecting procedures with common randomness and those without. Different from previous efforts, we study this problem from a quantizer design perspective. By analyzing an idealized setting, we provide an interpretation on the advantage of dithered quantization in the RDP setting, which further allows us to make a conceptual connection between randomized (dithered) quantizers and quantizers without common randomness. This new understanding leads to a new procedure for RDP coding based on multiple quantizers with offsets. Though the procedure can be viewed as intermediates between the two extremes, its explicit structure can be advantageous in some cases. Experimental results are given on both simple data sources and images to illustrate its behavior.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Theory (eg, control theory, learning theory, algorithmic game theory)