With 60% of the Internet traffic being video communications, a trend fueled by the mounting demand for remote presence, better video encoding techniques are in need. So far, deep learning has already proven capable to outperform traditional techniques at some image processing tasks. The Moving Picture, Audio and Data Coding by Artificial Intelligence (MPAI) organization has established a working group tasked with exploring AI-enhanced video coding. This project lies at the boundary between video coding and machine learning and aims at assessing to which extent a deep learning enhanced video codec can outperform state of the art video coding standards leveraging synergies with ongoing standardization efforts.

Contacts:
Attilio Fiandrotti
Project link: https://mpai.community