Dispider: Enabling Video LLMs with Active Real-Time Interaction via Disentangled Perception, Decision, and Reaction Paper • 2501.03218 • Published Jan 6, 2025 • 35
