SceneSmith
PulseAugur coverage of SceneSmith — every cluster mentioning SceneSmith across labs, papers, and developer communities, ranked by signal.
- 2026-07-13 product_launch MIT researchers developed SceneSmith, an AI system for generating 3D virtual environments for robot training. source
2 day(s) with sentiment data
-
SceneSmith uses AI agents to generate diverse 3D training scenes
SceneSmith has developed a system that utilizes three distinct agents—a designer, a critic, and an orchestrator—to generate diverse 3D training scenes. This approach addresses the previous bottleneck in sim-to-real tran…
-
MIT researchers use AI agents to generate 3D environments for robot training
Researchers at MIT have created SceneSmith, a system that uses AI agents to generate realistic 3D virtual environments for robot training. By employing GPT-5.2, specialized agents function as a designer, critic, and orc…
-
SceneSmith generates realistic indoor scenes for robot simulation
Researchers have developed SceneSmith, a novel agentic framework designed to generate realistic indoor environments for robot training simulations. This system uses a hierarchical approach with interacting VLM agents to…
-
New Benchmarking System Enhances Robotic Semantic Mapping
Researchers have developed OSMa-Bench++, an extension of OSMa-Bench, to create a more flexible and comprehensive benchmarking system for semantic mapping in robotics. This new framework utilizes prompt-generated synthet…