Open-World Neural Navigation
Real-time neural maps for unfamiliar indoor environments.
Four Papers at CVPR 2026
Four papers from our lab will be presented at CVPR 2026 in Nashville.
Highlights
See all
Compositional Vision-Language Reasoning
A training objective that improves compositional generalization.
Long-Form Video Memory Transformers
Memory-augmented transformers reasoning over hour-long videos.
Self-Supervised Medical Imaging
Cutting radiology annotation cost by 10x via contrastive pretraining.