Generative Image and Video Coding: Principles, Methods, and Emerging Directions

Speakers

• Jianhui Chang
  Cloud Computing Institute, China Telecom
• Giuseppe Valenzise
  Université Paris-Saclay, CNRS, CentraleSupélec, L2S, France
• Shiqi Wang
  Department of Computer Science, City University of Hong Kong

Tutorial outline

1. Foundations of Generative Visual Coding — Giuseppe Valenzise
   Coding paradigms, rate–distortion–perception, diffusion priors, and posterior reconstruction.
2. Diffusion Methods for Image and Video Coding — Jianhui Chang
   Bitstream representations, generation control, efficient decoding, and temporal modeling.
3. Generative Face Video Coding and Beyond — Shiqi Wang
   Compact facial representations, motion modeling, scalable coding, standardization, and future directions.

Citation and attribution

If you use the exposition, organization, or synthesis in these slides, please cite this tutorial. For specific methods, figures, and experimental results, please also cite the original publications credited in the slides. Third-party materials retain their original attribution. This citation refers to the tutorial slides; any future survey paper will have a separate bibliographic record.

Jianhui Chang, Giuseppe Valenzise, and Shiqi Wang. “Generative Image and Video Coding: Principles, Methods, and Emerging Directions.” Tutorial slides, IEEE International Conference on Image Processing (ICIP), 2026. Version 1.0.

@misc{chang2026generative,
  author       = {Chang, Jianhui and Valenzise, Giuseppe and Wang, Shiqi},
  title        = {Generative Image and Video Coding: Principles, Methods, and Emerging Directions},
  year         = {2026},
  howpublished = {Tutorial slides, IEEE International Conference on Image Processing (ICIP)},
  url          = {https://floatbutterfly.github.io/jianhui/tutorials/icip2026-gvc/},
  note         = {136 slides. Version 1.0}
}