A curated list of awesome test-time prior adaptation resources. Your contributions are always welcome!
-
CM-Bootstrap[Vucetic, Obradovic, Proc. ECML 2001] Classification on data with biased class distribution [PDF] [G-Scholar] -
MLLS[Latinne et al., Proc. ICML 2001] Adjusting the outputs of a classifier to new a priori probabilities may significantly improve classification accuracy: evidence from a multi-class problem in remote sensing [PDF] [G-Scholar] -
MLLS[Saerens et al., Neural Computation 2002] Adjusting the outputs of a classifier to new a priori probabilities: A simple procedure [PDF] [G-Scholar] -
...[Chan et al., Proc. ACL 2006] Estimating class priors in domain adaptation for word sense disambiguation [PDF] [G-Scholar] -
OEM[Yang and Zhou, Pattern Recognition 2008] Non-stationary data sequence classification using online class priors estimation [PDF] [G-Scholar] -
MLLS-PE[Du Plessis and Sugiyama, Neural Networks 2014] Semi-supervised learning of class balance under class-prior change by distribution matching [PDF] [G-Scholar] -
PTCA[Royer and Lampert, Proc. CVPR 2015] Classifier adaptation at prediction time [PDF] [G-Scholar] -
BBSE[Lipton et al., Proc. ICML 2018] Detecting and correcting for label shift with black box predictors [PDF] [G-Scholar] [CODE] -
RLLS[Azizzadenesheli et al., Proc. ICLR 2019] Regularized learning for domain adaptation under label shifts [PDF] [G-Scholar] [CODE] -
...[Sulc and Matas, Proc. ICCV Workshops 2019] Improving CNN classifiers by estimating test-time priors [PDF] [G-Scholar] -
BCTS[Alexandari et al., Proc. ICML 2020] Maximum likelihood with bias-corrected calibration is hard-to-beat at label shift adaptation [PDF] [G-Scholar] [CODE] -
MLLS-CM[Garg et al., Proc. NeurIPS 2020] A unified view of label shift estimation [PDF] [G-Scholar] -
...[Sulc et al., Proc. WACV 2020] Fungi recognition: A practical use case [PDF] [G-Scholar] [CODE] -
...[Šipka, Thesis 2021] Adaptation of CNN classifiers to prior shift [PDF] [G-Scholar] -
OGD[Wu et al., Proc. NeurIPS 2021] Online adaptation to label distribution shift [PDF] [G-Scholar] [CODE] -
...[Šipka et al., Proc. WACV 2022] The hitchhiker's guide to prior-shift adaptation [PDF] [G-Scholar] -
SADE[Zhang et al., Proc. NeurIPS 2022] Self-supervised aggregation of diverse experts for test-agnostic long-tailed recognition [PDF] [G-Scholar] [CODE] -
TTADC[Ma et al., Proc. MICCAI 2022] Test-time adaptation with calibration of medical image classification nets for label distribution shift [PDF] [G-Scholar] [CODE] -
DROPS[Wei et al., Proc. ICLR 2023] Distributionally robust post-hoc classifiers under prior shifts [PDF] [G-Scholar] [CODE] -
TTLSA[Sun et al., Proc. NeurIPS 2023] Beyond invariance: Test-time label-shift adaptation for distributions with" spurious" correlations [PDF] [G-Scholar] [CODE] -
...[Park et al., Proc. ICCV 2023] Label shift adapter for test-time adaptation under covariate and label shifts [PDF] [G-Scholar] -
FedCal[Xu and Huang, Proc. CIKM 2023] A joint training-calibration framework for test-time personalization with label shift in federated learning [PDF] [G-Scholar--] -
HANOL[Qian et al., Proc. ICDM 2023] Handling new class in online label shift [PDF] [G-Scholar] -
OLS-OFU[Wu et al., Proc. NeurIPS Workshops 2023] Online feature updates improve online (generalized) label shift adaptation [PDF] [G-Scholar] -
CPMKM[Wen et al., arXiv 2023] Class probability matching using kernel methods for label shift adaptation [PDF] [G-Scholar] -
...[Wei et al., Proc. ICML 2024] Learning label shift correction for test-agnostic long-tailed recognition [PDF] [G-Scholar] [CODE] -
Wav-O/-R[Qian et al., Proc. ICML 2024] Efficient non-stationary online learning by wavelets with applications to online distribution shift adaptation [PDF] [G-Scholar--] -
CPMCN[Wen et al., Proc. ICLR 2024] Class probability matching with calibrated networks for label shift adaption [PDF] [G-Scholar--] -
OLS-OFU[Wu et al., arXiv 2024] Online feature updates improve online (generalized) label shift adaptation [PDF] [G-Scholar]