memc
Technology:
Supplementary Material to Deep Contextual Video Compression
This document provides the supplementary material to our proposed deep contextual video compression (DCVC), including detailed network structures, training strategies, as well as additional experimental results to demonstrate the effectiveness of the proposed DCVC. The above part is the encoder and the below part is the decoder. For simplification, the entropy model is omitted. ResBlock represents plain residual block. The numbers represent channel dimensions.
Technology: