VidLLMs – Program

Workshop Program

Detailed schedule to be announced. Below is a tentative program structure.

Morning Session

8:30 AM – 8:40 AM Chairs' Opening Remarks

8:40 AM – 9:05 AM Invited Talk 1: Prof. Salman Khan
Where Video Reasoning Fails - and What to Do About It

9:05 AM – 9:30 AM Invited Talk 2: Prof. Lena Maier-Hein
Don't Lose the Sponge: Advancing Long-Context Video Understanding for Surgical Safety

9:30 AM – 9:55 AM Invited Talk 3: Dr. Ruben Villegas
Towards creating anything from any input: The Evolution of Multi-Modal Generative Models

9:55 AM – 10:20 AM Invited Talk 4: Dr. Afshin Dehghan
Flexible Length Video Tokenization

10:20 AM – 10:40 AM Break

10:40 AM – 11:05 AM Invited Talk 5: Prof. Mike Z Shou
Unified Multimodal Video Understanding and Generation

11:05 AM – 11:30 AM Invited Talk 6: Prof. Hilde Kuehne
Look, Listen, Adapt - Localization, Audio-Visual Reasoning, and Test-Time Adaptation for VideoLLMs

11:30 AM – 11:55 PM Invited Talk 7: Prof. Gerard Medioni
VideoLLMs in Prime Video

12:00 PM – 1:00 PM Lunch Break

Afternoon Session

1:00 PM – 2:15 PM Panel Discussion: Video LLMs and the Rise of World Models
Moderator: Prof. Rene Vidal
Panelists: Björn Ommer, Gerard Medioni, Octavia Camps and Hilde Kuehne

2:15 PM – 2:30 PM Break

2:30 PM – 2:45 PM Challenge Overview

2:45 PM – 3:00 PM Challenge Track 1 Winner Presentation
Reasoned-Aware Composed Video Retrieval (CoVR-R)

3:00 PM – 3:15 PM Challenge Track 2 Winner Presentation
See Beyond Frames: The Implicit Video Relational Reasoning (VRR) Challenge

3:15 PM – 3:30 PM Challenge Track 3 Winner Presentation
TimeLogic QA: Evaluating Temporal Reasoning in Videos

3:30 PM – 3:40 PM Poster Session Preparation

3:40 PM – 5:00 PM Poster Session