Archival paper
Long papers up to 8 pages and extended abstracts up to 4 pages (references and appendix not included), in NeurIPS 2026 style.
Machine Learning/CORE Unranked
PhysWorldAI 2026
The 1st Workshop on Physical World AI: Geometry, Characteristics, and Multimodal Sensing (NeurIPS 2026)
PhysWorldAI 2026 is the first workshop focused on Physical World AI, exploring how AI systems can perceive, represent, and reason about the physical world beyond visual appearance. It brings together researchers from computer vision, robotics, graphics, and multimodal learning to advance understanding in physical geometry, characteristics, and sensing, with a half-day program featuring invited talks, paper spotlights, posters, and live demos.
Key deadlines
Verified September 8, 2026Full paper
September 9, 2026
AoE
Workshop timeline
Submission and decisions
Full paperKey deadline
September 9, 2026 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Non-archival paper
Long papers up to 8 pages and extended abstracts up to 4 pages (references and appendix not included), in NeurIPS 2026 style.
Research areas in scope
01
Physical Geometry
3D/4D reconstructionarticulated and deformable scene understandinggeometry-aware world modelsphysically grounded view synthesis
02
Physical Characteristics
Material and physical property estimationmassfrictionstiffnesselasticitydeformabilityaffordancescontact-rich interactiondifferentiable simulationgenerative models of physical dynamics
03
Physical Sensors
Multimodal sensing and fusiontactileforce/torqueproprioceptiveRFaudiodepthIMUevent-based signals
04
Cross-cutting
Embodied world modelsrobot manipulationsim-to-real transfermultimodal simulatorsbenchmarksdatasetsevaluation protocolsresponsible deployment
Policies worth checking twice
- Submission format: Long papers up to 8 pages and extended abstracts up to 4 pages (references and appendix not included), in NeurIPS 2026 style.
- Review process: Double-blind review via OpenReview.
- Archival status: Both archival and non-archival papers are accepted; accepted papers will be posted on the workshop website.
- Dual submission: Concurrent submission to NeurIPS 2026 or other peer-reviewed venues is permitted if authors comply with those venues' policies.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 8, 2026