MM4SciReal:
Multimodal AI for Science and the Real World
The 1st Workshop, co-located with ACCV 2026
News
- Nov 4, 2026: The list of accepted papers is now available. Congratulations to all the authors!
- Oct 21, 2026: Author notifications for the archival and non-archival tracks have been sent. Archival camera-ready papers are due October 24, 2026 23:59 (AoE).
- Oct 2, 2026: Submissions are now closed for both tracks. Thank you to everyone who submitted!
- Sep 21, 2026: Invited speakers and the program have been announced.
- Aug 25, 2026: The Call for Papers is out.
- Jul 30, 2026: The workshop website is now live.
- Jul 27, 2026: MM4SciReal has been accepted as a workshop at ACCV 2026!
Overview
Scientific and real-world phenomena can only be captured through partial, noisy, and heterogeneous observations. This workshop aims to integrate these incomplete signals and connect them to scientific structure (physics, biology, causality) to recover underlying mechanisms, bringing together diverse research communities to shape the next decade of Science AI.
Topics of Interest
We welcome submissions on (but not limited to):
- Multimodal learning from incomplete, noisy, and heterogeneous scientific observations
- Integration of vision with language, biological and omics signals, sensor readings, molecular data, time series, and simulation outputs
- Learning methods informed by physics, chemistry, biology, geometry, or causal structure
- Multimodal foundation models for scientific discovery and real-world decision making
- Vision-language, audio-language, sensor-language, and 3D-language models
- Multimodal reasoning over images, text, tables, equations, graphs, videos, and simulations
- Data-efficient and label-efficient learning for scientific domains with limited or noisy annotations
- Applications in medicine, biology, materials science, robotics, earth and environmental science, and other scientific or real-world domains
See the full Call for Papers for submission details.
Important Dates
| Event | Date |
|---|---|
| Archival paper submission deadline | October 1, 2026 23:59 (AoE) |
| Non-archival submission deadline | October 1, 2026 23:59 (AoE) |
| Archival and non-archival author notification | October 21, 2026 |
| Archival camera-ready deadline | October 24, 2026 23:59 (AoE) |
| Workshop date | December 15, 2026 (Full-day) |
All deadlines are 23:59 Anywhere on Earth (AoE).
Venue
Room 1001, Osaka International Convention Center (Grand Cube Osaka), Osaka, Japan.
Sponsors
This workshop is supported by the following JST ASPIRE Projects:
- Deep Sensing: Jointly optimizing sensing and processing for CV/AI applications (Grant Number: JPMJAP2502)
- Formation of a Japan-Germany Multilayered R&D Network for Medical x AI and its Application to Industry Creation (Grant Number: JPMJAP2403)
Contact
Questions? Reach the organizers at mm4scireal@googlegroups.com.