Xiaolong Wang

A new method helps robots learn tasks using simulated data.

New methods enhance robots' multi-task learning and adaptability in complex environments.

A new system enhances robot navigation and object manipulation capabilities.

A method that combines language and physical properties for dynamic 3D scene creation.

This method simplifies creating new 3D views with limited images.

SpatialRGPT enhances object arrangement understanding in Vision Language Models.

New models improve image generation across various resolutions efficiently.

This article outlines a new approach using Test-Time Training for enhancing RNN performance.

A new platform enhances testing for autonomous vehicles through realistic racing simulations.

New methods improve robots' ability to grasp objects using 3D representations.

SPOT learns tasks by watching humans, making robots more helpful.

Researchers improve understanding of meson transitions and photon behavior.

Researchers are making robots capable of handling real-world chores effectively.

NaVILA helps robots navigate using language and vision.

A new model predicts hand movements from everyday language.