SeC: Advancing Complex Video Object Segmentation via Progressive Concept Construction Paper • 2507.15852 • Published Jul 21 • 38
BoostStep: Boosting mathematical capability of Large Language Models via improved single-step reasoning Paper • 2501.03226 • Published Jan 6 • 44
Dispider: Enabling Video LLMs with Active Real-Time Interaction via Disentangled Perception, Decision, and Reaction Paper • 2501.03218 • Published Jan 6 • 36
