BlinkMimic: Training a Humanoid Motion Tracker in Minutes on a Single GPU
cs.RO, cs.LG
Submitted: 2026-11-12
Updated: 2026-11-12
Code: https://github.com/fuyichen0725/blinkmimic
Terminology
Sources
- GMT: General Motion Tracking for Humanoid Whole-Body Control
- Adaptive Gradient Methods at the Edge of Stability
- Mastering Diverse Domains through World Models
- MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies
- BeyondMimic: From Motion Tracking to Versatile Humanoid Control via Guided Diffusion
- SuperMimic: Scaling Humanoid Whole-Body Control with One Differentiable World Model
- Muon is Scalable for LLM Training
- An Empirical Model of Large-Batch Training
- Gradients are Not All You Need
- SOAP: Improving and Stabilizing Shampoo using Adam
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation