Panoramic Multimodal Semantic Occupancy Prediction for Quadruped Robots

PanoMMOcc

PanoMMOcc is the first panoramic multimodal occupancy dataset for quadruped robots, featuring 360° panoramic, thermal, polarization, and LiDAR data, along with dense occupancy annotations collected under realistic robot motions and diverse scenes.

Demo

Code and Dataset

Coming soon!