Uncertainty Estimation for 3D Dense Prediction via Cross-Point Embeddings
This work addresses uncertainty estimation for 3D point cloud tasks, which is important for applications like autonomous driving and robotics, but it is incremental as it builds upon existing dense prediction pipelines.
The paper tackles the problem of uncertainty estimation for dense prediction tasks in 3D point clouds, introducing CUE and CUE+ methods that achieve well-calibrated uncertainty in geometric feature learning and reduce the Expected Calibration Error by 16.5% in semantic segmentation without compromising predictive performance.
Dense prediction tasks are common for 3D point clouds, but the uncertainties inherent in massive points and their embeddings have long been ignored. In this work, we present CUE, a novel uncertainty estimation method for dense prediction tasks in 3D point clouds. Inspired by metric learning, the key idea of CUE is to explore cross-point embeddings upon a conventional 3D dense prediction pipeline. Specifically, CUE involves building a probabilistic embedding model and then enforcing metric alignments of massive points in the embedding space. We also propose CUE+, which enhances CUE by explicitly modeling crosspoint dependencies in the covariance matrix. We demonstrate that both CUE and CUE+ are generic and effective for uncertainty estimation in 3D point clouds with two different tasks: (1) in 3D geometric feature learning we for the first time obtain wellcalibrated uncertainty, and (2) in semantic segmentation we reduce uncertainty's Expected Calibration Error of the state-of-the-arts by 16.5%. All uncertainties are estimated without compromising predictive performance.