ModelsSelected
A navigation model controls 4 robot types without fine-tuning
2000+真实场景搬进仿真！一个导航模型零样本“通吃”四种机器人本体
Why we picked it
For users needing to control multiple robot types, this model offers a unified solution without the need for separate training for each robot type.
Liangyuan Xinchuang released a new model trained on real-world scene data that can control different robot types (wheeled, tracked, legged, aerial) for navigation tasks without extra adjustments. The model leads in multiple robot benchmarks, including zero-shot transfer to specific robot bodies.
Original · 量子位
2000+真实场景搬进仿真！一个导航模型零样本“通吃”四种机器人本体
亮源新创的Physical Al路线清晰了
