Continual Learning for Table Detection in Document Images
The growing amount of data demands methods that can gradually learn from new samples. However, it is not trivial to continually train a network. Retraining a network with new data usually results in a phenomenon called “catastrophic forgetting”. In a nutshell, the performance of the model on the pre...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-09-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/12/18/8969 |
_version_ | 1797491593475784704 |
---|---|
author | Mohammad Minouei Khurram Azeem Hashmi Mohammad Reza Soheili Muhammad Zeshan Afzal Didier Stricker |
author_facet | Mohammad Minouei Khurram Azeem Hashmi Mohammad Reza Soheili Muhammad Zeshan Afzal Didier Stricker |
author_sort | Mohammad Minouei |
collection | DOAJ |
description | The growing amount of data demands methods that can gradually learn from new samples. However, it is not trivial to continually train a network. Retraining a network with new data usually results in a phenomenon called “catastrophic forgetting”. In a nutshell, the performance of the model on the previous data drops by learning from the new instances. This paper explores this issue in the table detection problem. While there are multiple datasets and sophisticated methods for table detection, the utilization of continual learning techniques in this domain has not been studied. We employed an effective technique called experience replay and performed extensive experiments on several datasets to investigate the effects of catastrophic forgetting. The results show that our proposed approach mitigates the performance drop by 15 percent. To the best of our knowledge, this is the first time that continual learning techniques have been adopted for table detection, and we hope this stands as a baseline for future research. |
first_indexed | 2024-03-10T00:50:34Z |
format | Article |
id | doaj.art-08e832142252498981048a3f3916bdec |
institution | Directory Open Access Journal |
issn | 2076-3417 |
language | English |
last_indexed | 2024-03-10T00:50:34Z |
publishDate | 2022-09-01 |
publisher | MDPI AG |
record_format | Article |
series | Applied Sciences |
spelling | doaj.art-08e832142252498981048a3f3916bdec2023-11-23T14:51:00ZengMDPI AGApplied Sciences2076-34172022-09-011218896910.3390/app12188969Continual Learning for Table Detection in Document ImagesMohammad Minouei0Khurram Azeem Hashmi1Mohammad Reza Soheili2Muhammad Zeshan Afzal3Didier Stricker4Department of Computer Science, Technical University of Kaiserslautern, 67663 Kaiserslautern, GermanyDepartment of Computer Science, Technical University of Kaiserslautern, 67663 Kaiserslautern, GermanyDepartment of Electrical and Computer Engineering, Kharazmi University, Tehran 1571914911, IranDepartment of Computer Science, Technical University of Kaiserslautern, 67663 Kaiserslautern, GermanyDepartment of Computer Science, Technical University of Kaiserslautern, 67663 Kaiserslautern, GermanyThe growing amount of data demands methods that can gradually learn from new samples. However, it is not trivial to continually train a network. Retraining a network with new data usually results in a phenomenon called “catastrophic forgetting”. In a nutshell, the performance of the model on the previous data drops by learning from the new instances. This paper explores this issue in the table detection problem. While there are multiple datasets and sophisticated methods for table detection, the utilization of continual learning techniques in this domain has not been studied. We employed an effective technique called experience replay and performed extensive experiments on several datasets to investigate the effects of catastrophic forgetting. The results show that our proposed approach mitigates the performance drop by 15 percent. To the best of our knowledge, this is the first time that continual learning techniques have been adopted for table detection, and we hope this stands as a baseline for future research.https://www.mdpi.com/2076-3417/12/18/8969table detectiondocument layout analysiscontinual learningincremental learningexperience replay |
spellingShingle | Mohammad Minouei Khurram Azeem Hashmi Mohammad Reza Soheili Muhammad Zeshan Afzal Didier Stricker Continual Learning for Table Detection in Document Images Applied Sciences table detection document layout analysis continual learning incremental learning experience replay |
title | Continual Learning for Table Detection in Document Images |
title_full | Continual Learning for Table Detection in Document Images |
title_fullStr | Continual Learning for Table Detection in Document Images |
title_full_unstemmed | Continual Learning for Table Detection in Document Images |
title_short | Continual Learning for Table Detection in Document Images |
title_sort | continual learning for table detection in document images |
topic | table detection document layout analysis continual learning incremental learning experience replay |
url | https://www.mdpi.com/2076-3417/12/18/8969 |
work_keys_str_mv | AT mohammadminouei continuallearningfortabledetectionindocumentimages AT khurramazeemhashmi continuallearningfortabledetectionindocumentimages AT mohammadrezasoheili continuallearningfortabledetectionindocumentimages AT muhammadzeshanafzal continuallearningfortabledetectionindocumentimages AT didierstricker continuallearningfortabledetectionindocumentimages |