
Glimpse from the Lab: VideoCartridges
Distilling spatiotemporal context into compact, reusable representations for efficient long-video understanding
AI Lab updates & research.

Distilling spatiotemporal context into compact, reusable representations for efficient long-video understanding

Data-free LoRA indexing and dynamic adaptation for open-vocabulary semantic segmentation under real-world domain shifts

Join us for high-impact research collaborations that target publication at top-tier conferences.

VocAlign: A framework for source-free domain adaptation in open-vocabulary segmentation

An approach to retrieve and merge LoRA adpapters for new domains
Browse other categories