Learning Deep Features for One-Class Classification

Pramuditha Perera, Vishal M. Patel
1/16/2018
cs.CV

Official code: PramuPerera/DeepOneClass (196 stars, LGPL-3.0, last commit 2019-03-26).

Abstract

We propose a deep learning-based solution for the problem of feature learning in one-class classification. The proposed method operates on top of a Convolutional Neural Network (CNN) of choice and produces descriptive features while maintaining a low intra-class variance in the feature space for the given class. For this purpose two loss functions, compactness loss and descriptiveness loss are proposed along with a parallel CNN architecture. A template matching-based framework is introduced to facilitate the testing process. Extensive experiments on publicly available anomaly detection, novelty detection and mobile active authentication datasets show that the proposed Deep One-Class (DOC) classification method achieves significant improvements over the state-of-the-art.

View on arXivView PDF

Code Implementations(1)

PramuPerera/DeepOneClassOfficial100%
LGPL-3.0

Deep learning based one class classification code targeting one class image classification. Tests carried out on Abnormal image detection, Novel image detection and Active Authentication reported state of the art results.

19646PythonJan 13, 20187 years agoLGPL-3.0

Similar Papers

Anomaly Detection using One-Class Neural Networks

82% match1 repo1672018

One-Class Convolutional Neural Network

82% match1 repo622019

Adversarially Learned One-Class Classifier for Novelty Detection

82% match1 repo2252018

Class-Specific Variational Auto-Encoder for Content-Based Image Retrieval

82% match2023

Exploring the Optimization Objective of One-Class Classification for Anomaly Detection

81% match2023

Learning Large Margin Sparse Embeddings for Open Set Medical Diagnosis

81% match2023

Cite this paper

@article{perera2018learning,
  title  = {Learning Deep Features for One-Class Classification},
  author = {Pramuditha Perera and Vishal M. Patel},
  year   = {2018},
  eprint = {1801.05365},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1801.05365}
}

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/1801.05365.svg)](https://codeofpaper.com/paper/1801.05365)

Discussion