AURA: Always-On Understanding and Real-Time Assistance via Video Streams
This addresses the need for real-time, interactive video understanding systems, which is an incremental improvement over existing streaming VideoLLMs.
The paper tackles the problem of enabling Video Large Language Models (VideoLLMs) to handle live video streams for continuous observation and timely response, achieving state-of-the-art performance on streaming benchmarks and supporting a real-time demo system running at 2 FPS on two 80G accelerators.
Video Large Language Models (VideoLLMs) have achieved strong performance on many video understanding tasks, but most existing systems remain offline and are not well-suited for live video streams that require continuous observation and timely response. Recent streaming VideoLLMs have made progress, yet current approaches often rely on decoupled trigger-response pipelines or are limited to captioning-style narration, reducing their effectiveness for open-ended question answering and long-horizon interaction. We propose AURA (Always-On Understanding and Real-Time Assistance), an end-to-end streaming visual interaction framework that enables a unified VideoLLM to continuously process video streams and support both real-time question answering and proactive responses. AURA integrates context management, data construction, training objectives, and deployment optimization for stable long-horizon streaming interaction. It achieves state-of-the-art performance on streaming benchmarks and supports a real-time demo system with ASR and TTS running at 2 FPS on two 80G accelerators. We release the AURA model together with a real-time inference framework to facilitate future research.