AI Paper Atlas

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

Beyond Global Alignment: Fine-Grained Motion-Language Retrieval via Pyramidal Shapley-Taylor Learning

Hanmo Chen, Guangtao Lyu, Chenghao Xu, Jiexi Yan, Xu Yang, Cheng Deng

ICML 2026 · Vision-language and multimodal › Human and embodied modeling · approach: Multimodal retrieval-augmented modeling · 0 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 STI distillation loss (lambda_S=1 vs 0) (ablation)PST · HumanML3D text-to-motionR@10: 39.12 -> 48.22Improved
increase lambda_S from 1 to 2 (ablation)PST · HumanML3D text-to-motionR@10: 48.22 -> 45.54Hurt
add joint-to-segment self-distillation loss (lambda_D=1 vs 0) (ablation)PST · HumanML3D text-to-motionR@10: 32.41 -> 48.22Improved
increase lambda_D from 1 to 3 (ablation)PST · HumanML3D text-to-motionR@10: 48.22 -> 35.49Hurt
lower compression ratio to 0.1 (from 0.25) (ablation)PST · HumanML3D text-to-motionR@10: 48.22 -> 31.67Hurt
raise compression ratio to 0.75 (from 0.25) (ablation)PST · HumanML3D text-to-motionR@10: 48.22 -> 31.76Hurt
use MotionPatch representation instead of Guo feature (ablation)PST · HumanML3D text-to-motionR@10: 30.14 -> 48.22Improved
add STI distillation loss (lambda_S=1 vs 0) (ablation)PST · KIT-ML text-to-motionR@10: 51.18 -> 57.87Improved
add self-distillation loss (lambda_D=1 vs 0) (ablation)PST · KIT-ML text-to-motionR@10: 42.10 -> 57.87Improved
increase lambda_S from 1 to 2 (ablation)PST · KIT-ML text-to-motionR@10: 57.87 -> 55.79No clear effect
replace Guo feature with MotionPatch representation (ablation)PST · KIT-ML text-to-motionR@10: 39.89 -> 57.87Improved
PST fine-grained pyramidal alignment versus prior best (Lyu et al.)HumanML3D All protocol · text-to-motionR@1: 11.80 -> 12.45Improved

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