AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

MotionGPT3: Human Motion as a Second Modality

Bingfan Zhu, Biao Jiang, Sunyi Wang, SHIXIANG TANG, Tao Chen, Linjie Luo, Youyi Zheng, Xin Chen

ICLR 2026 · Vision-language and multimodal › Human and embodied modeling · approach: Text-driven human motion generation · 39 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
use continuous VAE latents instead of VQ tokens (ablation)MotionGPT3 dual-stream · HumanML3D T2M (199 epochs)R-Precision Top-1: 0.525 vs. 0.258 on R@1Improved
use continuous VAE latents instead of VQ tokens (ablation)MotionGPT3 dual-stream · HumanML3D T2MMMDist: 2.667 vs. 5.364 on MMDistImproved
use continuous VAE latents instead of VQ tokens (ablation)MotionGPT3 dual-stream · HumanML3D M2T (99 epochs)R-Precision Top-3: 0.859 vs. 0.765 on R@3Improved
train VQ variant longer (399 epochs) (ablation)Bimodal+VQ · HumanML3D T2MFID: FID to 0.454, R@1 0.300 and MMDist 4.937Mixed
train VQ variant longer (199 epochs) (ablation)Bimodal+VQ · HumanML3D M2TR-Precision Top-3: R@3 dropping by 0.13 and BLEU@1 by 5.464Hurt
use dual-stream (bimodal) backbone instead of single-stream (ablation)MotionGPT3 with VAE · HumanML3D M2TR-Precision Top-3: 0.426 -> 0.853Improved
use dual-stream (bimodal) backbone instead of single-stream (ablation)MotionGPT3 with VQ · HumanML3D T2MR-Precision Top-3: 0.435 -> 0.532Improved
use dual-stream (bimodal) backbone instead of single-stream (ablation)MotionGPT3 with VAE · HumanML3D T2MFID: 0.489 -> 0.239Improved
add Stage II cross-modal alignment after Stage I (ablation)MotionGPT3 · HumanML3D T2MFID: 0.239 -> 0.215Improved
add Stage III joint fine-tuning (ablation)MotionGPT3 · HumanML3D M2TBleu@4: 18.328 -> 19.412Improved
add Stage I text-to-motion pretraining (ablation)MotionGPT3 · HumanML3D T2MFID: 0.325 -> 0.208Improved
increase text backbone from GPT-2 small to medium (ablation)MotionGPT3 · HumanML3D T2M (CFG 1.0)R-Precision Top-1: 0.534 -> 0.558 R@1, FID 0.222 -> 0.235Mixed

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