Nvidia wants to scale robot simulation training with Lyra 2.0

THE DECODER / 4/17/2026

📰 NewsSignals & Early TrendsTools & Practical UsageModels & Research

共有:

Key Points

Nvidia researchers introduced Lyra 2.0, which can generate large, coherent 3D environments from a single input photograph.
The generated 3D scenes are designed to be explored in real time, enabling faster inspection and iteration during content creation and testing.
Nvidia claims these image-to-3D environments can be used directly as inputs for robot simulation training, reducing manual environment setup.
The core goal is to scale robot simulation training by automating the creation of varied, usable virtual spaces from minimal real-world data.

Nvidia researchers have unveiled Lyra 2.0, a system that generates large, coherent 3D environments from a single photograph. The resulting scenes can be explored in real time and used directly in robot simulations.

The article Nvidia wants to scale robot simulation training with Lyra 2.0 appeared first on The Decoder.