# Hyperbolic Image Embeddings

## TL;DR

This paper explores the advantages of hyperbolic embeddings over traditional Euclidean and spherical embeddings in various computer vision tasks, showing they often yield better performance in practical scenarios.

## Contribution

It introduces the use of hyperbolic embeddings for computer vision tasks, highlighting their benefits over Euclidean and spherical methods.

## Key findings

- Hyperbolic embeddings outperform Euclidean and spherical embeddings in several tasks.
- Hyperbolic space better captures hierarchical and complex data structures.
- The study demonstrates practical scenarios where hyperbolic embeddings are advantageous.

## Abstract

Computer vision tasks such as image classification, image retrieval and few-shot learning are currently dominated by Euclidean and spherical embeddings, so that the final decisions about class belongings or the degree of similarity are made using linear hyperplanes, Euclidean distances, or spherical geodesic distances (cosine similarity). In this work, we demonstrate that in many practical scenarios hyperbolic embeddings provide a better alternative.