Home / AI Research / BLIP-Live: Balancing Timeliness & Firing Precision in Proactive Streaming Video Agents
BLAI Research
BLIP-Live: Balancing Timeliness & Firing Precision in Proactive Streaming Video Agents
by Salesforce · Kanchana Ranasinghe, Honglu Zhou, Dominick Reilly, Le Xue, Shusheng Yang, Juan Carlos Niebles
Streaming video understanding is becoming increasingly important for real-world interactive systems. Such systems must be proactive for natural and useful human interaction: they should respond as soon as sufficient evidence becomes available while avoiding premature, redundant, or incorrect responses. We present BLIP-Live, a framework unifying data, metrics, harness, and training around a shared formulation of proactivity as effective causal response decisions balancing the two competing objectives of responding vs silence. First, we construct BLIP-Live-780K dataset containing human annotations of proactive response content and timing, which defines this desired behavior. We then proposeā¦
Upvotes9
Comments2
Published15 Nov 2026