AI Paper Atlas

Research map / Video / Video generation

FineMotion: A Dataset and Benchmark with both Spatial and Temporal Annotation for Fine-grained Motion Generation and Editing

Bizhu Wu, Jinheng Xie, Meidan Ding, Zhe Kong, Jianfeng Ren, Ruibin Bai, Rong Qu, Linlin Shen

ICCV 2025 · Video › Video generation · approach: Text-to-motion diffusion · 6 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 fine-grained body-part text (DT) to training and inputT2M-GPT · FineMotion benchmarkFID: 0.123 -> 0.091Improved
add fine-grained body-part text (DT) to training and inputT2M-GPT · FineMotion benchmarkR-Precision Top-3: 0.781 -> 0.789Improved
add fine-grained body-part text (DT) to training and inputMDM · FineMotion benchmarkR-Precision Top-3: 0.606 -> 0.759Improved
add fine-grained body-part text (DT) to training and inputMDM · FineMotion benchmarkFID: 3.137 -> 0.488Improved
add fine-grained body-part text (DT) to training and inputMoMask · FineMotion benchmarkFID: 0.249 -> 0.055Improved
add fine-grained body-part text (DT) to training and inputMoMask · FineMotion benchmarkR-Precision Top-3: 0.753 -> 0.813Improved
encode coarse and detailed text separately (T&DT) rather than concatenated (TDT) (ablation)MoMask (BPMSD) · FineMotion benchmarkFID: 8.328 -> 0.088Improved
encode coarse and detailed text separately (T&DT) rather than concatenated (TDT) (ablation)MoMask (BPMP) · FineMotion benchmarkR-Precision Top-3: 0.628 -> 0.813Improved
use human-annotated DT in training instead of only automatic DT (ablation)T2M-GPT (BPMSD) · FineMotion benchmarkFID: 0.091 -> 0.091No clear effect
add detailed-text multitask training, evaluated with coarse text onlyT2M-GPT · HumanML3D-style T2M testFID: 0.123 -> 0.154Hurt
add detailed-text multitask training, evaluated with coarse text onlyMoMask · HumanML3D-style T2M testFID: 0.249 -> 0.120Improved
add detailed-text multitask training, evaluated with coarse text onlyMDM · HumanML3D-style T2M testR-Precision Top-3: 0.606 -> 0.746Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

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