AI Paper Atlas

Research map / Video / Video generation

GenM3: Generative Pretrained Multi-path Motion Model for Text Conditional Human Motion Generation

Junyu Shi, Lijiang Liu, Yong Sun, Zhiyuan Zhang, Jinni Zhou, Qiang Nie

ICCV 2025 · Video › Video generation · approach: Text-to-motion diffusion · 8 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 text pathway to motion-only multi-path transformer (ablation)GenM3 · HumanML3DFID: 0.058 -> 0.045Improved
add cross-modal pathway to motion-only multi-path transformer (ablation)GenM3 · HumanML3DFID: 0.058 -> 0.044Improved
add both text and cross-modal pathways to motion-only transformer (ablation)GenM3 · HumanML3DFID: 0.058 -> 0.035Improved
add text and cross-modal pathways (ablation)GenM3 · HumanML3DDiversity: 9.400 -> 9.341Hurt
use dense (fully activated) MoE instead of sparse MoE (ablation)GenM3 motion pathway · HumanML3DFID: 0.058 -> 0.046Improved
add multi-expert design to VQ (ablation)VQ-VAE · HumanML3DFID: 0.098 -> 0.048Improved
add multi-expert design to RVQ (ablation)RVQ · HumanML3DFID: 0.043 -> 0.032Improved
increase masked-decoding iterations from 1 to 10 (ablation)GenM3 · HumanML3DFID: 2.242 -> 0.046Improved
increase masked-decoding iterations from 10 to 15 (ablation)GenM3 · HumanML3DFID: 0.046 -> 0.060Hurt
increase pretraining data proportion from 0% to 100% (ablation)GenM3 · HumanML3DFID: 0.071 -> 0.046Improved
train text-conditional stage on all text-motion pairs (GenM3*) instead of HumanML3D onlyGenM3 · IDEA400 zero-shotFID: 4.430 -> 4.232Improved
train text-conditional stage on all text-motion pairs (GenM3*) instead of HumanML3D onlyGenM3 · HumanML3DFID: 0.046 -> 0.053Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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