Future Development Path of Intelligent Image and Video Coding

In recent years, intelligent image and video coding technologies have made significant progress. In terms of coding performance, they have approached or even surpassed classical image and video coding methods based on signal statistics, while enabling coding schemes optimized for both subjective quality and machine vision. However, their widespread adoption remains challenging due to high encoding/decoding complexity and the lack of dedicated hardware support. As the development of classical image and video coding methods has been stuck in a bottleneck, this article first examines the technical essence of intelligent image and video coding, then conducts an in-depth discussion focusing on the technological, standardization, and industrialization pathways of these coding technologies, aiming to promote further advancement in this field.