VRU_Basketball

This is a basketball dataset from AVS-VRU, which contains three dynamic basketball court scenes collected in Guangzhou (gz), Dongguan (dg), and Long360. It features large-scale motion, making it well-suited for evaluating dynamic scene reconstruction methods. Each scene contains 36 viewpoints, with videos captured at 1080p or 4K resolution for each viewpoint. The Guangzhou (gz) and Dongguan (dg) sequences are approximately 10 seconds long, each consisting of 250 frames at 1080p. The Long360… See the full description on the dataset page: Basketball.

Source: Hugging Face Hub (BestWJH/VRU_Basketball). Metadata imported from the dataset’s Hub tags.