2021-10: Rodrigo Marcuzzi wins Challenge at ICCV’21 Workshops on 4D LiDAR Panoptic Segmentation

The work “Contrastive Instance Association for 4D Panoptic Segmentation” by Rodrigo Marcuzzi et al. win the 4D LiDAR Panoptic Segmentation Challenge of the 6th ICCV Workshop on Benchmarking Multi-Target Tracking. The challenge was be based on the SemanticKITTI dataset, which is densely labeled in the spatial and temporal domain. The task was to assign a semantic and unique instance label to every 3D LiDAR point.