paper

Watch, Remember, Reason: Human-View Video Understanding with MLLMs

paperactiveprovisionalwatch-remember-reason-human-view-video-understanding-with-mllms-2429353e·1 events·first seen 9d ago

Aliases: Watch, Remember, Reason: Human-View Video Understanding with MLLMs

More like this (12)

Where Does the Answer Come From? Benchmarking View-Level Visual Evidence Identification in Multi-View MLLMs for Autonomous Driving Reasoning as Pattern Matching: Shared Mechanisms in Human and LLM Everyday Reasoning Scaling LLM Reasoning from Minimal Labels: A Semi-Supervised Framework with a Lightweight Verifier Multimodal Large Language Models visual language model Reasoning Language Models Multi-head Latent Attention (MLA)Learning from the Self-future: On-policy Self-distillation for dLLMs Reasoning in Memory (RiM)Latent World Recovery for Multimodal Learning with Missing Modalities Modeling Complex Behaviors: Multi-Personality Composition and Dynamic Switching in Vision-Language Models Gaze Heads: How VLMs Look at What They Describe

Recent events (1)

4arXiv · cs.AI·9d ago·source ↗

Survey: Human-View Video Understanding with MLLMs — Watch, Remember, Reason Framework

A new arXiv survey paper proposes a unified 'human-view' framework for analyzing multimodal LLM-based video understanding, organized around three functional abilities: watching (perception), remembering (memory), and reasoning. The authors introduce a formulation characterizing video understanding systems by perceptual representations, memory states, reasoning traces, and predictions, then survey methods, datasets, and benchmarks across these dimensions. The work covers challenges including spatio-temporal perception, long-video processing, streaming understanding, and faithful reasoning, with application domains spanning egocentric, sports, medical, and narrative video.

Long Context Evolution Multimodal Progress Watch, Remember, Reason: Human-View Video Understanding with MLLMs