Hyperbolic Image Segmentation
Mina Ghadimi Atigh, Julian Schoep, Erman Acar, Nanne van Noord, Pascal Mettes
20222022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)73 citationsDOIOpen Access PDF
Abstract
For image segmentation, the current standard is to perform pixel-level optimization and inference in Euclidean output embedding spaces through linear hyperplanes. In this work, we show that hyperbolic manifolds provide a valuable alternative for image segmentation and propose a tractable formulation of hierarchical pixel-level classification in hyperbolic space. Hyperbolic Image Segmentation opens up new possibilities and practical benefits for segmentation, such as uncertainty estimation and boundary information for free, zero-label generalization, and increased performance in low-dimensional output embeddings.
Topics & Concepts
Scale-space segmentationImage segmentationEmbeddingSegmentation-based object categorizationArtificial intelligenceSegmentationComputer scienceHyperplanePattern recognition (psychology)Image texturePixelBoundary (topology)Computer visionGeneralizationInferenceMathematicsMathematical analysisGeometryTopological and Geometric Data AnalysisMedical Image Segmentation TechniquesGenerative Adversarial Networks and Image Synthesis