SHTOcc: Effective 3D Occupancy Prediction with Sparse Head and Tail Voxels
This work solves performance and efficiency bottlenecks in 3D occupancy prediction for autonomous driving, representing an incremental advance with specific gains.
The paper tackles the problem of 3D occupancy prediction in autonomous driving by addressing voxel distribution issues, resulting in a 42.2% reduction in GPU memory usage, a 58.6% increase in inference speed, and a 7% accuracy improvement.
3D occupancy prediction has attracted much attention in the field of autonomous driving due to its powerful geometric perception and object recognition capabilities. However, existing methods have not explored the most essential distribution patterns of voxels, resulting in unsatisfactory results. This paper first explores the inter-class distribution and geometric distribution of voxels, thereby solving the long-tail problem caused by the inter-class distribution and the poor performance caused by the geometric distribution. Specifically, this paper proposes SHTOcc (Sparse Head-Tail Occupancy), which uses sparse head-tail voxel construction to accurately identify and balance key voxels in the head and tail classes, while using decoupled learning to reduce the model's bias towards the dominant (head) category and enhance the focus on the tail class. Experiments show that significant improvements have been made on multiple baselines: SHTOcc reduces GPU memory usage by 42.2%, increases inference speed by 58.6%, and improves accuracy by about 7%, verifying its effectiveness and efficiency. The code is available at https://github.com/ge95net/SHTOcc