We present StreamBridge, a simple yet effective framework that seamlessly transforms offline Video-LLMs into streaming-capable models. It addresses two fundamental challenges in adapting existing models into online scenarios: (1) limited capability for multi-turn real-time understanding, and (2) lack of proactive response mechanisms. Specifically, StreamBridge incorporates (1) a memory buffer combined with a round-decayed compression strategy, supporting long-context multi-turn interactions, and (2) a decoupled, lightweight activation model that can be effortlessly integrated into existing Video-LLMs, enabling continuous proactive responses. To further support StreamBridge, we construct Stream-IT, a large-scale dataset tailored for streaming video understanding, featuring interleaved video-text sequences and diverse instruction formats. Extensive experiments show that StreamBridge significantly improves the streaming understanding capabilities of offline Video-LLMs across various tasks, outperforming even proprietary models such as GPT-4o and Gemini 1.5 Pro. Simultaneously, it achieves competitive or superior performance on standard video understanding benchmarks.
View on arXiv@article{wang2025_2505.05467,
title={ StreamBridge: Turning Your Offline Video Large Language Model into a Proactive Streaming Assistant },
author={ Haibo Wang and Bo Feng and Zhengfeng Lai and Mingze Xu and Shiyu Li and Weifeng Ge and Afshin Dehghan and Meng Cao and Ping Huang },
journal={arXiv preprint arXiv:2505.05467},
year={ 2025 }
}