Splat2Act: From Captured Scenes to Personalized VLA Manipulation
Xiaomeng Ye, Krishna Jaganathan, Gyubeum Lim, Gunjan Chhablani, Muhammad Zubair Irshad, Zsolt Kira
THE PROBLEM
This paper focuses on Modern Robot LearningVision-Language-Action model (VLA)A model that takes images and language as input and outputs robot actions.. This paper shows how to take a pretrained Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. Manipulation & TasksManipulationUsing a robot arm or hand to move or interact with objects. Core ConceptsPolicyThe rule or model that maps observations or states to actions. and automatically adapt it to work in a specific user's home or office using just a phone scan—no real Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. demonstrations needed. It builds photorealistic 3D simulations of real spaces, generates 68,000+ synthetic Robot LearningTrainingThe process of fitting a model using data or experience. demos, and achieves 63.9% real-world pickup success on a Stretch Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions., enabling personalization without costly in-home data collection. Read the paper by tracking the Core ConceptsTaskThe job the robot is supposed to complete, such as pick-and-place, navigation, or drawer opening. definition, the Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. or data assumptions, and the evidence that supports the claimed improvement.
HOW IT WORKS
Task framing
Core method
Data and supervision
Evaluation evidence
KEY RESULTS
This paper shows how to take a pretrained Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. Manipulation & TasksManipulationUsing a robot arm or hand to move or interact with objects. Core ConceptsPolicyThe rule or model that maps observations or states to actions. and automatically adapt it to work in a specific user's home or office using just a phone scan—no real Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. demonstrations needed. It builds photorealistic 3D simulations of real spaces, generates 68,000+ synthetic Robot LearningTrainingThe process of fitting a model using data or experience. demos, and achieves 63.9% real-world pickup success on a Stretch Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions., enabling personalization without costly in-home data collection.
WHY DEVELOPERS SHOULD CARE
This paper shows how to take a pretrained Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. Manipulation & TasksManipulationUsing a robot arm or hand to move or interact with objects. Core ConceptsPolicyThe rule or model that maps observations or states to actions. and automatically adapt it to work in a specific user's home or office using just a phone scan—no real Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. demonstrations needed. It builds photorealistic 3D simulations of real spaces, generates 68,000+ synthetic Robot LearningTrainingThe process of fitting a model using data or experience. demos, and achieves 63.9% real-world pickup success on a Stretch Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions., enabling personalization without costly in-home data collection.
LIMITATIONS
The main limitation to check is whether the claimed behavior holds outside the paper's reported setup. That means testing across different Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. embodiments, scenes, objects, and data distributions.
WHAT COMES NEXT
The practical next step is independent reproduction with clear baselines, ablations, and stress tests. For a developer, the useful follow-up is to map the paper's Modern Robot LearningVision-Language-Action model (VLA)A model that takes images and language as input and outputs robot actions. assumptions onto a concrete Core ConceptsRobotA physical system with sensors and actuators that can observe the world and take actions. stack, then test the smallest version of the method that could run end to end.