ECCV 2026 MUCG workshop lineup: unified multimodal MLLMs and three new papers

mohitban47 · x · 2026-09-08

Researcher jmincho will attend ECCV 2026 in Malmö (Sep 8–11), keynoting the MUCG workshop on unified multimodal understanding and generation (Sep 8) and a workshop on curated data for efficient learning (Sep 9).

On Sep 11 he will present three posters: AnchorWeave (world-consistent video generation with retrieved local spatial memories), Imaginative Perception Tokens for spatial reasoning in multimodal LMs, and Physics Question Scene Graph for fine-grained evaluation of physical plausibility in text-to-video generation.

Related event: MUCG Workshop on Unified Multimodal Comprehension and Generation Lands at ECCV 2026(2 posts)→

Original post →

More from Multimodal

Multimodal channel →