Embodied AI Workshop at CVPR 2025

About the workshop

Host conference: The Conference on Computer Vision and Pattern Recognition (CVPR) (opens in new tab) | June 11-15, 2025

Panel speakers: Jianwei Yang

Workshop scientific advisor: Ade Famoti, Andrey Kolobov

Workshop organizer: Vivan Amin (opens in new tab), Jiaolong Yang (opens in new tab) (see all workshop organizers (opens in new tab))

The Embodied AI Workshop at CVPR 2025 (opens in new tab), will be held in conjunction with The Conference on Computer Vision and Pattern Recognition (CVPR) in Nashville, Tennessee. This year’s workshop focuses on the overarching theme of Real-World Applications: creating embodied AI solutions that are deployed in real-world environments, ideally in the service of real-world tasks. As embodied AI agents mature, the community is encouraged to promote work that transitions research from simulation and laboratory settings into practical, real-world applications.

This umbrella theme is divided into four topics:

• Embodied AI Solutions: As embodied AI solutions become more powerful, they should address more complex problems, particularly real-world challenges outside of simulation and the laboratory. While scientific advances are of interest, we actively seek work that applies embodied AI to real-world industry applications.
• Advances in Simulation: Advances in simulation have enabled many embodied AI algorithms. Procedural simulation, parameterized simulation, differentiable simulation, and world models are of interest, as are simulations based on the increasing numbers of large embodied datasets.
• Generative Methods for Embodied AI: Generative AI is becoming increasingly important for embodied artificial intelligence research. Topics such as generative AI for simulation, data generation, and policies (e.g., diffusion policies and world models) are of great interest.
• Foundation Models: Large-scale pretrained models adaptable to new tasks first emerged in language, speech, and vision domains. Increasingly, foundation models are being developed in robotics domains, including action, perception, problem-solving, and simulation. We invite research on adapting existing models to embodied problems and training embodied foundation models directly on such tasks.

Agenda

A detailed agenda will be posted on this page as soon as it’s available.

• Workshop talks: 8:50AM-5:30PM PT
• Poster session: 1:00PM-2:00PM PT

Challenges

The Embodied AI Workshop at CVPR 2025 will host several challenges:

• Social Mobile Manipulation Challenge: Developing embodied AI agents capable of performing long sequences of complex tasks through social interactions in dynamic, multi-agent environments.
• Multi-Object Rescue Challenge: Tasks involving reasoning about human intentions and planning within dynamic environments.
• Vision-Tactile Fusion Manipulation Challenge: Focusing on integrating vision and tactile signals for manipulation tasks.
• Open Vocabulary Mobile Manipulation Challenge: Encouraging agents to perform tasks using open vocabulary instructions.

Each challenge will have its own dataset, evaluation criteria, and submission guidelines. Winners will be announced during the workshop and may be invited to present their work.