AVGenL @ ECCV 2026 — Audio-Visual Generation & Learning Workshop

Overview

In this workshop, we aim to shine a spotlight on new approaches in audio-visual generation, as well as covering a wide range of topics related to audio-visual learning, where the convergence of auditory and visual signals unlocks new opportunities for advancing AI-driven creativity, understanding, and perception. We hope our workshop can bring together researchers, practitioners, and enthusiasts from diverse disciplines in both academia and industry to delve into the latest developments, challenges, and breakthroughs in audio-visual generation and learning. Some of the main topics that will be covered are:

Audio-Visual Generation: sounding video generation, audio-driven talking head animation, video-to-audio generation, etc.
Audio-Visual Foundation Models: large-scale audio-visual pre-training, foundational video/audio generation and understanding models, etc.
Real-World Applications: deepfake detection, speech recognition, speech enhancement, video dubbing, audio-visual localization, etc.
Ethics and Safety: navigating the implications of synthetic media and increasingly capable machine perception.

Call for Papers

Submissions

We're inviting submissions related to audio-visual generation and learning, or any of the topics above. If you're interested in giving a 5-minute talk about your work and/or presenting a poster, please submit an extended abstract via our Google Form below. Think of this as a way to showcase your newly published or ongoing research efforts and get feedback and ideas from other researchers in the field.

We invite submissions for work that has already been accepted in other conferences/journals (e.g., CVPR, ICCV, ECCV, ICML, ICLR, NeurIPS, ICASSP, Interspeech, IJCV, TPAMI, TMLR, etc.), as well as new, work-in-progress research efforts/ideas.

Authors should use the official ECCV 2026 author kit listed at https://eccv.ecva.net/Conferences/2026/SubmissionPolicies. Please ensure that you use the template in camera-ready format with all authors and affiliations included. Accepted papers will be posted on this website as-is. In the Overleaf template, you can search for the comments saying "TODO FINAL" in main.tex to find what needs to be updated.

The submitted papers must be at most 4 pages, excluding references. Note that we also welcome shorter (e.g., 2-page) extended abstracts.

Accepted presenters will have their paper posted on our website, will present a short 5-minute talk at the workshop, and may also present a poster during our poster session. Since they are extended abstracts, they will not be included in the main conference proceedings.