AI Paper Atlas

Research map / Video / Video-language understanding

MotionLLM: Understanding Human Behaviors from Human Motions and Videos

Ling-Hao Chen, Shunlin Lu, Ailing Zeng, Hao Zhang, Benyou Wang, Ruimao Zhang, Lei Zhang

TPAMI 2025 · Video › Video-language understanding · approach: Event and anomaly reasoning in video · 116 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add video data with unpaired instruction data to motion model (ablation)MotionLLM · MoVid-Bench motion ablationaccuracy (All): 38.48 -> 48.07Improved
add paired Motion-XQA instruction tuning on top of unpaired data (motion) (ablation)MotionLLM · MoVid-Bench motion ablationaccuracy (All): 48.07 -> 49.50Improved
add paired Motion-XQA (sequential perception aspect) (ablation)MotionLLM · MoVid-Bench motion ablationsequential accuracy: 46.20 -> 36.84Hurt
add motion data with unpaired instruction data to video model (ablation)MotionLLM · MoVid-Bench video ablationaccuracy (All): 42.53 -> 41.49No clear effect
add paired Motion-XQA instruction tuning on top of unpaired data (video) (ablation)MotionLLM · MoVid-Bench video ablationaccuracy (All): 41.49 -> 48.94Improved
joint motion-video training instead of text-only baseline for motion (ablation)MotionLLM · MoVid-Bench motionaccuracy (All): 38.48 -> 49.50Improved
fine-tune on BABEL-QA for exact string matching (ablation)MotionLLM · BABEL-QA testoverall accuracy: 0.372 -> 0.436Improved
MotionLLM versus MotionGPT on motion benchmarkMotionLLM · MoVid-Bench motionaccuracy (All): 36.86 -> 49.50Improved
MotionLLM versus Video-LLaVA on video benchmarkMotionLLM · MoVid-Bench videoaccuracy (All): 42.53 -> 49.00Improved
MotionLLM versus VideoChat on zero-shot MVBenchMotionLLM · MVBench (7 sub-tasks)average accuracy: 30.1 -> 31.6Improved
MotionLLM versus Video-chat2 on zero-shot ActivityNet-QAMotionLLM · ActivityNet-QAaccuracy: 49.1 -> 53.3Improved

Later papers that compared with this method

6 later accepted papers report MotionLLM* (base) as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy