Bridging Modalities, Spanning Time: Structured Memory for Ultra-Long Agentic Video Reasoning

• URL: http://arxiv.org/abs/2605.08271v1
• Date: Fri, 08 May 2026 03:21:47 GMT
• Title: Bridging Modalities, Spanning Time: Structured Memory for Ultra-Long Agentic Video Reasoning
• Abstract summary: MAGIC-Video is a framework built around a multimodal memory graph with interleaved narrative chain.<n>On EgoLifeQA, Ego-R1 and MM-Lifelong, MAGIC-Video consistently outperforms strong general-purpose, long-video, and agentic baselines.
• Score: 82.97398529552166
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Understanding ultra-long videos such as egocentric recordings, live streams, or surveillance footage spanning days to weeks, remains a challenge. For current multimodal LLMs: even with million-token context windows, frame budgets cover only tens of minutes of densely sampled video, and most evidence is discarded before inference begins. Memory-augmented and agentic approaches help with scale, but their retrieval remains fragmented across modalities and lacks long-range narrative summaries that span days or weeks. We propose \textbf{MAGIC-Video}, a training-free framework built around a multimodal memory graph with interleaved narrative chain: the graph unifies episodic, semantic, and visual content through six typed edges and supports cross-modal retrieval, while the chain distils long-horizon entity biographies and recurring activity events. At inference time, an agentic loop interleaves graph retrieval with narrative fact injection, covering both the modality and time dimensions of ultra-long video in a single retrieval pipeline. On EgoLifeQA, Ego-R1 and MM-Lifelong, MAGIC-Video consistently outperforms strong general-purpose, long-video, and agentic baselines, with gains of 10.1, 7.4, and 5.9 points over the prior best agentic system on each benchmark. Code is available at https://github.com/lijiazheng0917/MAGIC-video.

Related papers

• VideoScout: Learning Agentic Active Exploration with Adaptive Reasoning Pacing for Long Video Understanding [61.00809248898572]
  Long video understanding is a problem in which an agent reads the video turn by turn along the temporal axis, deciding at each turn how fast to watch, what evidence to retain, when to revisit uncertain segments, and when to stop and answer.<n>We propose textbfVideoScout, a multi-turn reasoning agent that instantiates the SEA paradigm through adaptive reasoning pacing.<n>Experiments on long video understanding and reasoning benchmarks demonstrate that our 7B model achieves strong performance compared with existing trained 7B agentic models.
  arXiv  Detail & Related papers  (2026-09-14T14:04:03Z)
• Online Video Agent Harness for Long Video Understanding [35.044765942623876]
  Long video understanding often behaves like a visual needle-in-a-haystack problem.<n>Existing video agents often rely on query-agnostic offline preprocessing or ad hoc tool sets.<n>We present VideoXAgent, a purely online video-agent harness for long video understanding.
  arXiv  Detail & Related papers  (2026-09-11T13:13:03Z)
• Graph it first! Enabling Reasoning on Long-form Egocentric Videos through Scene Graphs [12.707862116540914]
  We introduce a framework for egocentric video question answering (VQA) that overcomes these input constraints through Egocentric Scene Graphs (EgoSGs)<n>By representing videos as compact, text-based scene graphs, our method preserves the essential visual and temporal information of the original video in a symbolic form.<n>On HD-EPIC VQA, our method achieves state-of-the-art results, outperforming strong video-based baselines on multiple models.
  arXiv  Detail & Related papers  (2026-06-24T13:55:45Z)
• Agentic Very Long Video Understanding [39.34545320553102]
  EGAgent is an enhanced agentic framework centered on entity scene graphs, which represent people, places, objects, and their relationships over time.<n>Our system equips a planning agent with tools for structured search and reasoning over these graphs, as well as hybrid visual and audio search capabilities, enabling detailed, cross-modal, and temporally coherent reasoning.<n>EgoLifeQA and Video-MME (Long) datasets show that our method achieves state-of-the-art performance on EgoLifeQA (57.5%) and competitive performance on Video-MME (Long) (74.1%) for complex longitudinal video understanding tasks.
  arXiv  Detail & Related papers  (2026-01-26T05:20:47Z)
• WorldMM: Dynamic Multimodal Memory Agent for Long Video Reasoning [66.24870234484668]
  We introduce WorldMM, a novel multimodal memory agent that constructs and retrieves from multiple complementary memories.<n>WorldMM significantly outperforms existing baselines across five long video question-answering benchmarks.
  arXiv  Detail & Related papers  (2025-12-02T05:14:52Z)
• LongVT: Incentivizing "Thinking with Long Videos" via Native Tool Calling [87.98096428508181]
  LongVT is an end-to-end agentic framework that enables "Thinking with Long Videos" via interleaved Multimodal Chain-of-Tool-Thought.<n>We exploit LMMs' inherent temporal grounding ability as a native video cropping tool to zoom in on a specific video clip and resample finer-grained video frames.<n>Our training dataset consists of 247.9K samples for tool-integrated cold-start supervised fine-tuning, 1.6K samples for agentic reinforcement learning, and 15.4K samples for agentic reinforcement fine-tuning.
  arXiv  Detail & Related papers  (2025-11-25T19:22:48Z)
• Mixture of Contexts for Long Video Generation [72.96361488755986]
  We recast long-context video generation as an internal information retrieval task.<n>We propose a simple, learnable sparse attention routing module, Mixture of Contexts (MoC), as an effective long-term memory retrieval engine.<n>As we scale the data and gradually sparsify the routing, the model allocates compute to salient history, preserving identities, actions, and scenes over minutes of content.
  arXiv  Detail & Related papers  (2025-08-28T17:57:55Z)
• Episodic Memory Representation for Long-form Video Understanding [52.33907540905242]
  Large Video Language Models excel at general video understanding but struggle with long-form context window limits.<n>We introduce Video-EM, a training free framework inspired by the principles of human memory.<n>Video-EM achieves performance gains of 4-9 percent over respective baselines while utilizing fewer frames.
  arXiv  Detail & Related papers  (2025-08-13T04:33:07Z)

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.