AI 导读
首个版本支持 3 类具身场景下的 6 个基准: ✈️ 空中导航 🏠 室内导航 🎯 移动目标跟踪 研究者可用 SimpleNav 复现实验、开发新的 VLA 模型,并探索跨数据集与跨场景泛化。 🔗 Github: https://t.co/j56PYtEvqv 🎥 看看 SimpleNav 的实际效果👇
正文
The first release supports 6 benchmarks across 3 embodied scenarios:
✈️ Aerial navigation
🏠 Indoor navigation
🎯 Mobile object tracking
Researchers can use SimpleNav to reproduce experiments, develop new VLA models, and explore cross-dataset and cross-scenario generalization.
🔗 Github: https://t.co/j56PYtEvqv
🎥 Check out SimpleNav in action👇
来源:@OpenBMB · x.com