Research map / Video / Video generation
HGM³: Hierarchical Generative Masked Motion Modeling with Hard Token Mining
Minjae Jeong, Yechan Hwang, Jaejin Lee, Sungyoon Jung, Won Hwa Kim
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes HGM³, a hierarchical generative masked motion model with hard token mining for text-to-motion generation, addressing text ambiguity and motion dynamics complexity.
Problem
Text-to-motion generation is challenging due to ambiguity in text descriptions and the complexity of human motion dynamics, which limit masked motion models.
Approach
Integrates a hierarchical generative masked motion modeling framework with hard token mining so that difficult motion tokens receive focused training; the abstract describes two key components, with details not available in the provided text.
Experimental setup
Builds on: masked motion models
Closest papers by meaning
- MMM: Generative Masked Motion Model (CVPR 2024 · similarity 0.93)
- MoMask: Generative Masked Modeling of 3D Human Motions (CVPR 2024 · similarity 0.90)
- SnapMoGen: Human Motion Generation from Expressive Texts (NeurIPS 2025 · similarity 0.90)
- Scaling Large Motion Models with Million-Level Human Motions (ICML 2025 · similarity 0.89)
- OMG: Towards Open-vocabulary Motion Generation via Mixture of Controllers (CVPR 2024 · similarity 0.89)
- ReMoGPT: Part-Level Retrieval-Augmented Motion-Language Models (AAAI 2025 · similarity 0.88)
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.
