Think While Watching: Online Streaming Segment-Level Memory for Multi-Turn Video Reasoning in Multimodal Large Language Models

• URL: http://arxiv.org/abs/2603.11896v1
• Date: Thu, 12 Mar 2026 13:13:50 GMT
• Title: Think While Watching: Online Streaming Segment-Level Memory for Multi-Turn Video Reasoning in Multimodal Large Language Models
• Abstract summary: Think While Watching is a memory-anchored streaming video reasoning framework.<n>It preserves continuous segment-level memory during multi-turn interaction.<n>It improves single-round accuracy by 2.6% on StreamingBench and by 3.79% on OVO-Bench.
• Score: 25.48403877647779
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Multimodal large language models (MLLMs) have shown strong performance on offline video understanding, but most are limited to offline inference or have weak online reasoning, making multi-turn interaction over continuously arriving video streams difficult. Existing streaming methods typically use an interleaved perception-generation paradigm, which prevents concurrent perception and generation and leads to early memory decay as streams grow, hurting long-range dependency modeling. We propose Think While Watching, a memory-anchored streaming video reasoning framework that preserves continuous segment-level memory during multi-turn interaction. We build a three-stage, multi-round chain-of-thought dataset and adopt a stage-matched training strategy, while enforcing strict causality through a segment-level streaming causal mask and streaming positional encoding. During inference, we introduce an efficient pipeline that overlaps watching and thinking and adaptively selects the best attention backend. Under both single-round and multi-round streaming input protocols, our method achieves strong results. Built on Qwen3-VL, it improves single-round accuracy by 2.6% on StreamingBench and by 3.79% on OVO-Bench. In the multi-round setting, it maintains performance while reducing output tokens by 56%. Code is available at: https://github.com/wl666hhh/Think_While_Watching/

Related papers

• StreamArena: Toward Continuous, Interactive, and Long-Horizon Agentic Streaming Video Understanding [42.88155526332792]
  We introduce StreamArena, a benchmark for hour-scale, interactive streaming video understanding.<n> evaluation across diverse systems exposes a tension between continuous interaction and long-horizon multimodal comprehension.<n>We address this tension with StreamMind, a two-tier architecture that assigns latency-critical interaction and proactive monitoring to independently scheduled workers.
  arXiv  Detail & Related papers  (2026-08-06T07:46:37Z)
• LiveStarPro: Proactive Streaming Video Understanding with Hierarchical Memory for Long-Horizon Streams [59.485485426790966]
  We introduce LiveStarPro, a live streaming assistant that is designed for proactive video understanding over long-horizon streams.<n>LiveStarPro consistently surpasses existing methods, attaining a 28.9% improvement in semantic correctness.<n>Its streaming key-value cache further yields a 1.58x inference speedup over the same model without caching.
  arXiv  Detail & Related papers  (2026-06-16T11:18:05Z)
• Harnessing Streaming Video in the Wild [53.23721420272668]
  Vision-Language Models (VLMs) are increasingly required to process video streams in applications such as video-call assistants, live commentary, and embodied robots.<n>An ideal streaming system should support proactive interaction, long-horizon memory, and real-time processing.<n>ExistingVLMs excel at offline video understanding but fall short in streaming capabilities and lack dedicated infrastructure for streaming deployment.
  arXiv  Detail & Related papers  (2026-06-07T13:00:19Z)
• Don't Pause: Streaming Video-Language Synchrony for Online Video Understanding [69.296913137409]
  We introduce a novel paradigm for online video understanding: Streaming Video-Language Synchrony (SVLS)<n>We present LyraV, a live streaming assistant built upon a hierarchical control framework with two core innovations.<n>First, the Frame-Driven Transition Controller (FDTC), makes high-level semantic decisions on when to continue speaking, start a new response, or stay silent.<n>Second, the Streaming Token Pacer (SToP), a plug-and-play lightweight predictive module, dynamically adapts the language generation rate to match the pace of the visual content.
  arXiv  Detail & Related papers  (2026-06-05T07:29:20Z)
• Thinking in Streaming Video [30.61790766076081]
  ThinkStream is a framework for streaming video reasoning based on a Watch--Think--Speak paradigm.<n> Reasoning-Compressed Streaming Memory (RCSM) treats intermediate reasoning traces as compact semantic memory.<n>Experiments on multiple streaming video benchmarks show that ThinkStream significantly outperforms existing online video models.
  arXiv  Detail & Related papers  (2026-03-13T12:33:36Z)
• Video Streaming Thinking: VideoLLMs Can Watch and Think Simultaneously [69.0264594684213]
  Video Streaming Thinking (VST) is a novel paradigm for streaming video understanding.<n>It supports a thinking while watching mechanism, which activates reasoning over incoming video clips during streaming.<n>VST improves timely comprehension and coherent cognition while preserving real-time responsiveness.
  arXiv  Detail & Related papers  (2026-03-12T17:59:51Z)
• Think-as-You-See: Streaming Chain-of-Thought Reasoning for Large Vision-Language Models [14.21980212001207]
  Motivated by the streaming nature of video data, we investigate two streaming reasoning paradigms for LVLMs.<n>To better match streaming inputs, we propose textbfThink-as-You-See (TaYS), a unified framework enabling true concurrent reasoning.
  arXiv  Detail & Related papers  (2026-03-03T11:24:55Z)
• Event-VStream: Event-Driven Real-Time Understanding for Long Video Streams [11.495597616926274]
  Event-VStream represents continuous video as a sequence of discrete, semantically coherent events.<n>System detects meaningful state transitions by integrating motion, semantic, and predictive cues.<n>System maintains around 70% GPT-5 win rate on 2-hour Ego4D streams.
  arXiv  Detail & Related papers  (2026-01-22T05:05:53Z)
• Speak While Watching: Unleashing TRUE Real-Time Video Understanding Capability of Multimodal Large Language Models [15.345320064963575]
  Multimodal Large Language Models (MLLMs) have achieved strong performance across many tasks, yet most systems remain limited to offline inference.<n>Recent streaming methods reduce latency by interleaving perception and generation, but still enforce a sequential perception-generation cycle.<n>We propose a parallel streaming framework that relaxes positional continuity through three designs: Overlapped, Group-Decoupled, and Gap-Isolated.
  arXiv  Detail & Related papers  (2026-01-11T10:12:11Z)
• StreamingAssistant: Efficient Visual Token Pruning for Accelerating Online Video Understanding [29.539015046656615]
  We propose token pruning as a means to reduce context length while retaining critical information.<n>Specifically, we introduce a novel redundancy metric, Maximum Similarity to Spatially Adjacent Video Tokens (MSSAVT)<n>We also design a masked pruning strategy that ensures only mutually unadjacent tokens are pruned.
  arXiv  Detail & Related papers  (2025-12-14T05:35:11Z)
• video-SALMONN S: Streaming Audio-Visual LLMs Beyond Length Limits via Memory [51.03819128505358]
  Video-SALMONN S is first to process 3-hour videos at 1 FPS and 360p resolution under a fixed memory budget.<n>A test-time-training memory module continually updates token representations to capture long-range dependencies.<n>A prompt-dependent memory reader retrieves context-relevant content from fixed-size memory.
  arXiv  Detail & Related papers  (2025-10-13T08:20:15Z)
• Streaming Video Understanding and Multi-round Interaction with Memory-enhanced Knowledge [57.01131456894516]
  Current video understanding models struggle with processing long video sequences, supporting multi-turn dialogues, and adapting to real-world dynamic scenarios.<n>We propose StreamChat, a training-free framework for streaming video reasoning and conversational interaction.<n>Our framework incorporates a parallel system scheduling strategy that enhances processing speed and reduces latency, ensuring robust performance in real-world applications.
  arXiv  Detail & Related papers  (2025-01-23T08:33:10Z)
• InternLM-XComposer2.5-OmniLive: A Comprehensive Multimodal System for Long-term Streaming Video and Audio Interactions [104.90258030688256]
  This project introduces disentangled streaming perception, reasoning, and memory mechanisms, enabling real-time interaction with streaming video and audio input.<n>This project simulates human-like cognition, enabling multimodal large language models to provide continuous and adaptive service over time.
  arXiv  Detail & Related papers  (2024-12-12T18:58:30Z)
• Flash-VStream: Memory-Based Real-Time Understanding for Long Video Streams [78.72965584414368]
  We present Flash-VStream, a video-language model that simulates the memory mechanism of human. Compared to existing models, Flash-VStream achieves significant reductions in latency inference and VRAM consumption. We propose VStream-QA, a novel question answering benchmark specifically designed for online video streaming understanding.
  arXiv  Detail & Related papers  (2024-06-12T11:07:55Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.