Zhiqi Zhang

Adding just 5% of target-embodiment data during pretraining can boost transfer performance by over 13 percentage points, highlighting the critical role of embodiment exposure in VLA models.

The challenge revealed that innovative techniques can significantly enhance visibility for autonomous vehicles in rainy conditions, a key hurdle for safe navigation.