Point Cloud Semantic Segmentation with Sparse and Inhomogeneous Annotations
This work addresses the challenge of handling non-uniform sparse annotations in real-world weakly supervised point cloud segmentation, offering a robust solution for applications like autonomous driving or robotics.
The paper tackles the problem of point cloud semantic segmentation with sparse and inhomogeneous annotations by proposing an Adaptive Annotation Distribution Network (AADNet), which achieves comprehensive performance improvements at multiple label rates and distributions without prior restrictions.
Utilizing uniformly distributed sparse annotations, weakly supervised learning alleviates the heavy reliance on fine-grained annotations in point cloud semantic segmentation tasks. However, few works discuss the inhomogeneity of sparse annotations, albeit it is common in real-world scenarios. Therefore, this work introduces the probability density function into the gradient sampling approximation method to qualitatively analyze the impact of annotation sparsity and inhomogeneity under weakly supervised learning. Based on our analysis, we propose an Adaptive Annotation Distribution Network (AADNet) capable of robust learning on arbitrarily distributed sparse annotations. Specifically, we propose a label-aware point cloud downsampling strategy to increase the proportion of annotations involved in the training stage. Furthermore, we design the multiplicative dynamic entropy as the gradient calibration function to mitigate the gradient bias caused by non-uniformly distributed sparse annotations and explicitly reduce the epistemic uncertainty. Without any prior restrictions and additional information, our proposed method achieves comprehensive performance improvements at multiple label rates and different annotation distributions.