Can contrastive learning avoid shortcut solutions?
Joshua Robinson, Li Sun, Ke Yu, Kayhan Batmanghelich, Stefanie Jegelka, Suvrit Sra
Abstract
(IFM), a method for altering positive and negative samples in order to guide contrastive models towards capturing a wider variety of predictive features. Empirically, we observe that IFM reduces feature suppression, and as a result improves performance on vision and medical imaging tasks. The code is available at: https://github.com/joshr17/IFM.
Topics & Concepts
GeneralizationComputer scienceTask (project management)Feature (linguistics)Code (set theory)Representation (politics)Artificial intelligenceVariety (cybernetics)Feature learningFeature extractionPattern recognition (psychology)Machine learningNatural language processingMathematicsLinguisticsEconomicsMathematical analysisPolitical scienceSet (abstract data type)ManagementLawProgramming languagePoliticsPhilosophyDomain Adaptation and Few-Shot LearningAdvanced Image and Video Retrieval TechniquesAI in cancer detection