FOUND: Foundation Data for Industrial Tech Transfer @ ECCV 2026

We are happy to announce that our workshop proposal for the 2nd FOUND Workshop, “FOUND: Foundation Data for Industrial Tech Transfer @ ECCV 2026”, has been accepted! The workshop will be held on the morning of September 9, 2026, in Malmö, Sweden, in conjunction with ECCV 2026.

This workshop follows the FOUND Workshop accepted at ICCV 2025 and continues the discussion on Foundation Data: domain-grounded data designed to bridge academic progress in foundation models and practical impact in real-world deployments.

About FOUND Workshop

Foundation models have pushed computer vision forward, but their reliability often drops in real-world, domain-specific deployments due to long-tail distributions, domain shift, and operational constraints such as privacy, safety, and cost. Closing this gap requires not only better models, but also domain-tailored datasets that capture rare yet critical corner cases and realistic variability.

Building on the first FOUND workshop, we refer to this missing piece as Foundation Data: fine-grained, domain-grounded data designed to bridge academic progress and practical impact. FOUND is organized around two coupled pillars, Foundation Data creation and practical tech transfer, with a key focus on Physical AI and World Models.

Topics of Interest

The workshop focuses on the following topics:

• Foundation Data creation: collection, curation, and labeling for non-Internet and domain-specific settings
• Self-supervised, semi-supervised, and weakly-supervised learning for domain-grounded data
• Synthetic and simulation data for bridging domain gaps and covering rare corner cases
• Robustness under domain shift, out-of-distribution conditions, and long-tail distributions in industrial deployments
• Physical AI and World Models: data and transfer strategies for embodied and interactive systems
• Deployment-aligned evaluation protocols and responsible data governance for real-world AI

Keynote Speakers

The ECCV 2026 edition will feature the following keynote speakers:

• Iro Armeni, Stanford University
• Mark Pollefeys, ETH Zurich / Microsoft Mixed Reality and AI Lab
• Amir Bar, Imperial College London / AMI Labs
• Anastasia Varava, SEBx / Link

Call for Nominations

Unlike a standard peer-reviewed call for papers, this workshop uses a self-nomination and curatorial invitation format. Nominated work is reviewed openly by the organizing committee for topical fit, potential to spark discussion, and overall balance of the poster session.

Selected work will be presented at the workshop poster session, with at least one presenter attending ECCV 2026 in person. Invitations do not come with official publication or Springer proceedings. Only presentation titles and presenter names will appear on the workshop website; abstracts and posters stay within the workshop audience.

The nomination form is open until August 19, 2026, 23:59 AoE.