ProtoPShare
Dawid Rymarczyk, Łukasz Struski, Jacek Tabor, Bartosz Zieliński
Abstract
In this work, we introduce an extension to ProtoPNet called ProtoPShare which shares prototypical parts between classes. To obtain prototype sharing we prune prototypical parts using a novel data-dependent similarity. Our approach substantially reduces the number of prototypes needed to preserve baseline accuracy and finds prototypical similarities between classes. We show the effectiveness of ProtoPShare on the CUB-200-2011 and the Stanford Cars datasets and confirm the semantic consistency of its prototypical parts in user-study.
Topics & Concepts
Computer scienceConsistency (knowledge bases)Extension (predicate logic)Baseline (sea)Similarity (geometry)Information retrievalSemantic similarityData miningArtificial intelligenceProgramming languageImage (mathematics)GeologyOceanographyAdvanced Neural Network ApplicationsExplainable Artificial Intelligence (XAI)Machine Learning and Data Classification