OccGen: Generative Multi-modal 3D Occupancy Prediction for Autonomous Driving

arXiv:2404.15014v1 Announce Type: new
Abstract: Existing solutions for 3D semantic occupancy prediction typically treat the task as a one-shot 3D voxel-wise segmentation perception problem. These discriminative methods focus on learning the mapping between the inputs and occupancy map in a single step, lacking the ability to gradually refine the occupancy map and the reasonable scene imaginative capacity to complete the local regions somewhere. In this paper, we introduce OccGen, a simple yet powerful generative perception model for the task of …

abstract arxiv autonomous autonomous driving cs.cv driving focus generative inputs map mapping modal multi-modal perception prediction refine segmentation semantic solutions type voxel wise