CVPR 2026 · Workshop

Area Chair
Workshop 2026

June 2, 2026Room 601/603, CVPR 2026

A gathering of the CVPR area chair community. Town halls, policy, new frontiers in computer vision, lightning talks, posters.

Program

Schedule

08:30
Breakfast
09:00
Introduction
09:30
Town Hall
10:30
Coffee break
11:00
Lightning talks
12:15
Lunch
13:00
Panel discussion · Past and future CVPR / ICCV program committees
14:30
Lightning talks (continued)
16:00
Coffee break
16:30
Town Hall
17:30
Poster session
18:30
Dinner
Lightning Talks

A series of short research talks.

11:00 Morning Session

  • Anand BhattadPrincipia: Do Video Models Understand Physical Reality? A Relational Evaluation
  • Mohamed ElhoseinyTowards Scalable and Structured Understanding in Video LLMs
  • Guy GilboaHow to Encode World Knowledge?
  • Ismini LourentzouGenerative Foundations for Embodied AI
  • Sagie Benaim3D-Aware Latent Tokens for 3D Reconstruction, Generation, and Understanding
  • Adam KortylewskiFrom 2D Representation Learning to Understanding the 3D World
  • Huaizu JiangUniCorrn: Unified Correspondence Transformer Across 2D and 3D
  • Eunbyung ParkScalable Feed-forward 3D Gaussian Splatting
  • Liwei WangLearning from Videos for 3D Spatial Understanding and Interaction
  • Roni SenguptaFrom Pixels to Physics: Understanding the Physical World from Visual Data
  • Davide MoltisantiVideo understanding: from fine-grained recognition to interpretable approaches
  • Michael RyooMotion World Models for Robot Control

14:30 Afternoon Session

  • Wayne WuFrom Scaling up to Scaling out: Reality World Simulators for Physical AI
  • Shubham TulsianiDo we need geometry for 3D, or 3D for robotics?
  • Liang ZhengT2I Pretraining: Problems, What's working, and What's next
  • Raymond A. YehComputer Vision Beyond Task Performance
  • Ziwei LiuFrom Egocentric Perception to Embodied Intelligence: Building the World in First Person
  • Andre AraujoTIPSv2: Advancing Vision-Language Pretraining with Enhanced Patch-Text Alignment
  • Renjie LiaoFlowing Through Time: Advancing Motion Forecasting and Training-Free Visual Editing
  • Liangyan GuiTowards Scalable and Autonomous Humanoid Whole-Body Loco-Manipulation
  • Stephan AlanizGenerative Explainable AI
  • Xiangyu YueOneThinker: All-in-one Reasoning Model for Image and Video
  • Zhuang LiuBuilding Fully Open Multimodal Foundation Models: From Reasoning to Generation
  • Cherry (Nanxuan) ZhaoUnified controllable content creation models
  • Deepti GhadiyaramRobust and Efficient Multimodal Intelligence
  • Or LitanyBending Generative Models Without Retraining
  • Chris ThomasStructuring Meaning and Model Behavior for Trustworthy Multimodal AI
Posters

Bring your work.

Poster spec
The poster boards are the same as those used in the main conference. Your poster should not exceed 213 cm × 107 cm.

Organizers

Angela Dai

Technical University of Munich

Adriana Kovashka

University of Pittsburgh

Chen Change Loy

Nanyang Technological University

Vladimir Pavlovic

Rutgers University

Alexander G. Schwing

University of Illinois Urbana-Champaign

Shaoting Zhang

Shanghai Jiao Tong University