OccFeat: Self-supervised Occupancy Feature Prediction for Pretraining BEV Segmentation Networks

arXiv:2404.14027v1 Announce Type: cross
Abstract: We introduce a self-supervised pretraining method, called OcFeat, for camera-only Bird's-Eye-View (BEV) segmentation networks. With OccFeat, we pretrain a BEV network via occupancy prediction and feature distillation tasks. Occupancy prediction provides a 3D geometric understanding of the scene to the model. However, the geometry learned is class-agnostic. Hence, we add semantic information to the model in the 3D space through distillation from a self-supervised pretrained image foundation model. Models pretrained with our method exhibit improved …

abstract arxiv bird cs.cv cs.lg distillation feature geometry however network networks prediction pretraining segmentation tasks type understanding via view