CONTINUAL LEARNING THROUGH ONE-CLASS CLASSIFICATION USING VAE

被引：0

作者：

Wiewel, Felix ^{[1
]}

Brendle, Andreas ^{[1
]}

Yang, Bin ^{[1
]}

机构：

[1] Univ Stuttgart, Inst Signal Proc & Syst Theory, Stuttgart, Germany

来源：

2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING | 2020年

关键词：

Continual Learning; One-class Classification; Variational Autoencoder; Deep Generative Replay;

D O I：

10.1109/icassp40776.2020.9054743

中图分类号：

O42 [声学];

学科分类号：

070206 ; 082403 ;

摘要：

Artificial neural networks (ANNs) suffer from catastrophic forgetting, a sharp decrease in performance on previously learned tasks, when trained on a new task without constant rehearsal. In this paper, we propose a new method for overcoming this phenomenon based on one-class classification. It is not only able to incrementally learn new but also detect unknown classes. This is a desirable property, since it enables the detection of new and unknown classes in a stream of data and adaption to a changing environment. Experiments on commonly used continual learning setups show competitive results and verify the concept.

引用

页码：3307 / 3311

页数：5

共 50 条

[1] Active Learning for One-Class Classification Using Two One-Class Classifiers
Schlachter, Patrick
Yang, Bin
[J]. 2018 26TH EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO), 2018, : 1197 - 1201
[2] ROBUST AND UNCERTAINTY-AWARE VAE (RU-VAE) FOR ONE-CLASS CLASSIFICATION
Sharma, Renuka
Awate, Suyash P.
[J]. 2022 IEEE INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING (IEEE ISBI 2022), 2022,
[3] Active Learning for One-Class Classification
Barnabe-Lortie, Vincent
Bellinger, Colin
Japkowicz, Nathalie
[J]. 2015 IEEE 14TH INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND APPLICATIONS (ICMLA), 2015, : 390 - 395
[4] A Semi-supervised Generalized VAE Framework for Abnormality Detection using One-Class Classification
Sharma, Renuka
Mashkaria, Satvik
Awate, Suyash P.
[J]. 2022 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2022), 2022, : 1302 - 1310
[5] Document Classification with One-class Multiview Learning
Chen, Bin
Li, Bin
Pan, Zhisong
Feng, Aimin
[J]. 2009 INTERNATIONAL CONFERENCE ON INDUSTRIAL AND INFORMATION SYSTEMS, PROCEEDINGS, 2009, : 289 - +
[6] One-Class Classification with Extreme Learning Machine
Leng, Qian
Qi, Honggang
Miao, Jun
Zhu, Wentao
Su, Guiping
[J]. MATHEMATICAL PROBLEMS IN ENGINEERING, 2015, 2015
[7] Kernel Matrix Learning for One-Class Classification
Wang, Chengqun
Lu, Jiangang
Hu, Chonghai
Sun, Youxian
[J]. ADVANCES IN NEURAL NETWORKS - ISNN 2008, PT I, PROCEEDINGS, 2008, 5263 : 753 - +
[8] Learning Deep Features for One-Class Classification
Perera, Pramuditha
Patel, Vishal M.
[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2019, 28 (11) : 5450 - 5463
[9] Improving Active Learning for One-Class Classification Using Dimensionality Reduction
Ghazel, Mohsen
Japkowicz, Nathalie
[J]. ADVANCES IN ARTIFICIAL INTELLIGENCE, CANADIAN AI 2017, 2017, 10233 : 39 - 44
[10] Resampling approach for one-Class classification; Resampling approach for one-Class classification
Lee, Hae-Hwan
Park, Seunghwan
Im, Jongho
[J]. Pattern Recognition, 2023, 143

← 1 2 3 4 5 →