Bharatesh Chakravarthi

Scaling up vision-language models and carefully adapting them for extreme far-distance video ReID dramatically improves performance, suggesting that pre-trained knowledge can overcome significant degradation from distance and viewpoint changes.

Event cameras can rescue robot navigation in low-light environments where RGB fails, as demonstrated by a new multimodal policy that leverages event data for robust imitation learning.

Despite excelling at abstract inference, today's best video language models still struggle with basic visual grounding in real-world urban traffic scenarios, highlighting a critical perception-reasoning gap.