1
Geometrically Constrained Outlier Synthesis
ICML 2026新研究，利用几何约束合成异常值，为异常检测提供高质量训练样本。
arXiv:2603.08413v2 Announce Type: replace-cross Abstract: Deep neural networks for image classification often exhibit overconfidence on out-of-distrib…
ICML 2026新研究，利用几何约束合成异常值，为异常检测提供高质量训练样本。
arXiv:2603.08413v2 Announce Type: replace-cross Abstract: Deep neural networks for image classification often exhibit overconfidence on out-of-distrib…