Deep Anomaly Detection with Outlier Exposure (ICLR 2019)
Distributionally robust neural networks for group shifts
The main features of uoguelph-mlrg/confidence_estimation are: Long-Tailed / Out-of-Distribution Recognition.
Open-source alternatives to uoguelph-mlrg/confidence_estimation include: alinlab/confident_classifier. facebookresearch/invariantriskminimization. hendrycks/outlier-exposure — Deep Anomaly Detection with Outlier Exposure (ICLR 2019). kohpangwei/group_dro — Distributionally robust neural networks for group shifts. shiyuliang/odin-pytorch — Principled Detection of Out-of-Distribution Examples in Neural Networks. ufoym/imbalanced-dataset-sampler — A (PyTorch) imbalanced dataset sampler for oversampling low frequent classes and undersampling high frequent ones.