Equivariant Covariance Tensors: Guaranteed SPD Uncertainty for Tensor-Valued Geometric Learning
等变协方差张量:张量值几何学习的保证SPD不确定性
机构 * College of Intelligent Robotics and Advanced Manufacturing, Fudan University(复旦大学智能机器人与先进制造学院) ; School of Microelectronics, Fudan University(复旦大学微电子学院) ; ByteDance(字节跳动) ; School of Information Science and Engineering, East China University of Science and Technology(华东理工大学信息科学与工程学院)
AI总结 针对几何深度学习中张量值预测的不确定性量化难题,提出E(3)等变UQ框架,通过分解协方差与矩阵指数法保证正定性,构建LE-ESO损失,在两类数据集上验证了方法的竞争力与不确定性估计的有效性。
Comments Accepted to ICML 2026