Learning Spread-out Local Feature Descriptors

Xu Zhang, Felix X. Yu, Sanjiv Kumar, Shih-Fu Chang
8/21/2017
cs.CV

Official code: ColumbiaDVMM/Spread-out_Local_Feature_Descriptor (41 stars, last commit 2019-02-20).

Abstract

We propose a simple, yet powerful regularization technique that can be used to significantly improve both the pairwise and triplet losses in learning local feature descriptors. The idea is that in order to fully utilize the expressive power of the descriptor space, good local feature descriptors should be sufficiently "spread-out" over the space. In this work, we propose a regularization term to maximize the spread in feature descriptor inspired by the property of uniform distribution. We show that the proposed regularization with triplet loss outperforms existing Euclidean distance based descriptor learning techniques by a large margin. As an extension, the proposed regularization technique can also be used to improve image-level deep feature embedding.

View on arXivView PDF

Code Implementations(1)

ColumbiaDVMM/Spread-out_Local_Feature_DescriptorOfficial100%

Code for Learning Spread-out Local Feature Descriptors

4114PythonJul 24, 20177 years ago

Similar Papers

Working hard to know your neighbor's margins: Local descriptor learning loss

86% match1 repo5342017

Residual Learning for Image Point Descriptors

83% match2023

Learning Local Image Descriptors with Deep Siamese and Triplet Convolutional Networks by Minimising Global Loss Functions

82% match1 repo342015

Improving the HardNet Descriptor

81% match1 repo262020

Learning Structured Representations by Embedding Class Hierarchy with Fast Optimal Transport

80% match1 repo32024

Domain Generalization via Discrete Codebook Learning

79% match2025

Cite this paper

@article{zhang2017learning,
  title  = {Learning Spread-out Local Feature Descriptors},
  author = {Xu Zhang and Felix X. Yu and Sanjiv Kumar and Shih-Fu Chang},
  year   = {2017},
  eprint = {1708.06320},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1708.06320}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1708.06320.svg)](https://codeofpaper.com/paper/1708.06320)

Discussion