AI Paper Atlas

Research map / Video / Video generation

Next-Scale Autoregressive Models for Text-to-Motion Generation

Zhiwei Zheng, Shibo Jin, Lingjie Liu, Mingmin Zhao

CVPR 2026 · Video › Video generation · approach: Compositional text-to-video diffusion · 3 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 cross-scale hierarchical refinement (ablation)MoScale · HumanML3D testTop-1: 0.481 -> 0.534Improved
add cross-scale hierarchical refinement (ablation)MoScale · HumanML3D testFID: 0.176 -> 0.090Improved
add in-scale temporal refinement (ablation)MoScale base · HumanML3D testTop-1: 0.481 -> 0.497Improved
add temporal refinement on top of hierarchical refinement (ablation)MoScale · HumanML3D testFID: 0.090 -> 0.046Improved
add temporal refinement on top of hierarchical refinement (ablation)MoScale · HumanML3D testTop-1: 0.534 -> 0.540Improved
add temporal refinement only (ablation)MoScale base · HumanML3D high-complexity descriptionsVLM text-alignment score: 1.89 -> 1.92No clear effect
use ordered coarse-to-fine generation instead of undirected mask-and-repredict (ablation)MoScale · HumanML3D testTop-1: 0.507 -> 0.540Improved
raise max corruption rate from 0.2 to 0.6 (ablation)hierarchical refinement · HumanML3D testFID: 0.096 -> 0.046Improved
raise max corruption rate from 0.6 to 0.8 (ablation)hierarchical refinement · HumanML3D testFID: 0.046 -> 0.074Hurt
increase transformer depth from 4 to 16 layers (ablation)MoScale · HumanML3D testFID: 0.096 -> 0.046Improved
increase refinement iterations beyond (1,2,5,10) (ablation)MoScale · HumanML3D testTop-1: 0.540 -> 0.539No clear effect
increase refinement iterations from (1,1,1,1) to (1,2,5,10) (ablation)MoScale · HumanML3D testTop-1: 0.535 -> 0.540Improved

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