Tag: vision-language-action
Browse our exclusive articles!
Boost Reinforcement Learning with Vision-Language-Action
Enhance robotic RL efficiency using Vision-Language-Action Jump-Starting for faster learning and better real-world task performance.
StarVLA: Modular Codebase for Vision-Language-Action Models
Discover StarVLA, an open-source modular codebase for Vision-Language-Action models enabling flexible research, reproducibility, and benchmark integration.
DIAL: Decoupling Intent and Action for Advanced VLA Models
Discover DIAL, a novel framework that enhances Vision-Language-Action models by separating intent and action via latent world modeling for robust robot con...
ETA-VLA: Efficient Token Adaptation for Vision-Language-Action Models
Discover ETA-VLA, a method that reduces computation in Vision-Language-Action models by 61% while maintaining 94% accuracy for autonomous driving.
Enhancing Robot Policy Robustness with Q-DIG Prompt Generation
Discover how Q-DIG improves Vision-Language-Action models by generating diverse prompts to boost robot policy robustness and identify failure modes.