Introducing Urdu Digits Dataset with Demonstration of an Efficient and Robust Noisy Decoder-Based Pseudo Example Generator
In the present work, we propose a novel method utilizing only a decoder for generation of pseudo-examples, which has shown great success in image classification tasks. The proposed method is particularly constructive when the data are in a limited quantity used for semi-supervised learning (SSL) or...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-09-01
|
| Series: | Symmetry |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-8994/14/10/1976 |
| _version_ | 1797469836383617024 |
|---|---|
| author | Wisal Khan Kislay Raj Teerath Kumar Arunabha M. Roy Bin Luo |
| author_facet | Wisal Khan Kislay Raj Teerath Kumar Arunabha M. Roy Bin Luo |
| author_sort | Wisal Khan |
| collection | DOAJ |
| description | In the present work, we propose a novel method utilizing only a decoder for generation of pseudo-examples, which has shown great success in image classification tasks. The proposed method is particularly constructive when the data are in a limited quantity used for semi-supervised learning (SSL) or few-shot learning (FSL). While most of the previous works have used an autoencoder to improve the classification performance for SSL, using a single autoencoder may generate confusing pseudo-examples that could degrade the classifier’s performance. On the other hand, various models that utilize encoder–decoder architecture for sample generation can significantly increase computational overhead. To address the issues mentioned above, we propose an efficient means of generating pseudo-examples by using only the generator (decoder) network separately for each class that has shown to be effective for both SSL and FSL. In our approach, the decoder is trained for each class sample using random noise, and multiple samples are generated using the trained decoder. Our generator-based approach outperforms previous state-of-the-art SSL and FSL approaches. In addition, we released the Urdu digits dataset consisting of 10,000 images, including 8000 training and 2000 test images collected through three different methods for purposes of diversity. Furthermore, we explored the effectiveness of our proposed method on the Urdu digits dataset by using both SSL and FSL, which demonstrated improvement of 3.04% and 1.50% in terms of average accuracy, respectively, illustrating the superiority of the proposed method compared to the current state-of-the-art models. |
| first_indexed | 2024-03-09T19:26:41Z |
| format | Article |
| id | doaj.art-480ec671f11346f49c0e70f39dd7fa38 |
| institution | Directory Open Access Journal |
| issn | 2073-8994 |
| language | English |
| last_indexed | 2024-03-09T19:26:41Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Symmetry |
| spelling | doaj.art-480ec671f11346f49c0e70f39dd7fa382023-11-24T02:49:58ZengMDPI AGSymmetry2073-89942022-09-011410197610.3390/sym14101976Introducing Urdu Digits Dataset with Demonstration of an Efficient and Robust Noisy Decoder-Based Pseudo Example GeneratorWisal Khan0Kislay Raj1Teerath Kumar2Arunabha M. Roy3Bin Luo4School of Computer and Technology, Anhui University, Hefei 230039, ChinaSchool of Computing, Dublin City University, SFI for Research Training in Artificial Intelligence, Dublin 9, IrelandDepartment of Software Engineering, School of Computing, National University of Computer and Emerging Sciences, Islamabad 44000, PakistanAerospace Engineering Department, University of Michigan, Ann Arbor, MI 48109, USASchool of Computer and Technology, Anhui University, Hefei 230039, ChinaIn the present work, we propose a novel method utilizing only a decoder for generation of pseudo-examples, which has shown great success in image classification tasks. The proposed method is particularly constructive when the data are in a limited quantity used for semi-supervised learning (SSL) or few-shot learning (FSL). While most of the previous works have used an autoencoder to improve the classification performance for SSL, using a single autoencoder may generate confusing pseudo-examples that could degrade the classifier’s performance. On the other hand, various models that utilize encoder–decoder architecture for sample generation can significantly increase computational overhead. To address the issues mentioned above, we propose an efficient means of generating pseudo-examples by using only the generator (decoder) network separately for each class that has shown to be effective for both SSL and FSL. In our approach, the decoder is trained for each class sample using random noise, and multiple samples are generated using the trained decoder. Our generator-based approach outperforms previous state-of-the-art SSL and FSL approaches. In addition, we released the Urdu digits dataset consisting of 10,000 images, including 8000 training and 2000 test images collected through three different methods for purposes of diversity. Furthermore, we explored the effectiveness of our proposed method on the Urdu digits dataset by using both SSL and FSL, which demonstrated improvement of 3.04% and 1.50% in terms of average accuracy, respectively, illustrating the superiority of the proposed method compared to the current state-of-the-art models.https://www.mdpi.com/2073-8994/14/10/1976semi-supervised learning (SSL)few-shot learning (FSL)encoder–decoderUrdu digits datasetdeep learning |
| spellingShingle | Wisal Khan Kislay Raj Teerath Kumar Arunabha M. Roy Bin Luo Introducing Urdu Digits Dataset with Demonstration of an Efficient and Robust Noisy Decoder-Based Pseudo Example Generator Symmetry semi-supervised learning (SSL) few-shot learning (FSL) encoder–decoder Urdu digits dataset deep learning |
| title | Introducing Urdu Digits Dataset with Demonstration of an Efficient and Robust Noisy Decoder-Based Pseudo Example Generator |
| title_full | Introducing Urdu Digits Dataset with Demonstration of an Efficient and Robust Noisy Decoder-Based Pseudo Example Generator |
| title_fullStr | Introducing Urdu Digits Dataset with Demonstration of an Efficient and Robust Noisy Decoder-Based Pseudo Example Generator |
| title_full_unstemmed | Introducing Urdu Digits Dataset with Demonstration of an Efficient and Robust Noisy Decoder-Based Pseudo Example Generator |
| title_short | Introducing Urdu Digits Dataset with Demonstration of an Efficient and Robust Noisy Decoder-Based Pseudo Example Generator |
| title_sort | introducing urdu digits dataset with demonstration of an efficient and robust noisy decoder based pseudo example generator |
| topic | semi-supervised learning (SSL) few-shot learning (FSL) encoder–decoder Urdu digits dataset deep learning |
| url | https://www.mdpi.com/2073-8994/14/10/1976 |
| work_keys_str_mv | AT wisalkhan introducingurdudigitsdatasetwithdemonstrationofanefficientandrobustnoisydecoderbasedpseudoexamplegenerator AT kislayraj introducingurdudigitsdatasetwithdemonstrationofanefficientandrobustnoisydecoderbasedpseudoexamplegenerator AT teerathkumar introducingurdudigitsdatasetwithdemonstrationofanefficientandrobustnoisydecoderbasedpseudoexamplegenerator AT arunabhamroy introducingurdudigitsdatasetwithdemonstrationofanefficientandrobustnoisydecoderbasedpseudoexamplegenerator AT binluo introducingurdudigitsdatasetwithdemonstrationofanefficientandrobustnoisydecoderbasedpseudoexamplegenerator |