Image Clustering: Utilizing Teacher-Student Model and Autoencoder
Complete and high-quality labeled dataset is indispensable for image classification. Considering the often arduous task of data labeling, clustering algorithms are commonly utilized in the preliminary stages of data classification to perform preliminary categorization. Traditional clustering algorit...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10255740/ |
| _version_ | 1797668059762130944 |
|---|---|
| author | Zhengshun Fei Haibo Gong Junhao Guo Jinglong Wang Wuyin Jin Xinjian Xiang Xiasheng Ding Ni Zhang |
| author_facet | Zhengshun Fei Haibo Gong Junhao Guo Jinglong Wang Wuyin Jin Xinjian Xiang Xiasheng Ding Ni Zhang |
| author_sort | Zhengshun Fei |
| collection | DOAJ |
| description | Complete and high-quality labeled dataset is indispensable for image classification. Considering the often arduous task of data labeling, clustering algorithms are commonly utilized in the preliminary stages of data classification to perform preliminary categorization. Traditional clustering algorithms, such as K-means and Gaussian Mixture Model, often struggle to effectively cluster images. In this paper, we propose a novel image clustering method utilizing teacher-student model and autoencoder. Specifically, the teacher model is a fully connected autoencoder and the student model consists of multiple convolutional autoencoders. We firstly obtain sub-datasets by applying the K-means algorithm to cluster the output of the teacher model, then utilize the student model to achieve clustering through an iterative data exchange method where the same data can converge. In addition, the proposed method can filter low-quality data to a certain extent by recording the frequency of exchanges because clustering is achieved during multiple exchanges of data. Experiments on a synthetic dataset and three benchmark image datasets (MNIST, FashionMNIST and USPS) show that our method can achieve satisfactory clustering results. |
| first_indexed | 2024-03-11T20:23:36Z |
| format | Article |
| id | doaj.art-b5ce4f89c24f46309a11123f4fd8fa25 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-11T20:23:36Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-b5ce4f89c24f46309a11123f4fd8fa252023-10-02T23:00:23ZengIEEEIEEE Access2169-35362023-01-011110484610485710.1109/ACCESS.2023.331728210255740Image Clustering: Utilizing Teacher-Student Model and AutoencoderZhengshun Fei0https://orcid.org/0000-0001-8111-690XHaibo Gong1https://orcid.org/0009-0009-0830-765XJunhao Guo2Jinglong Wang3https://orcid.org/0000-0002-7129-649XWuyin Jin4https://orcid.org/0000-0003-3741-3610Xinjian Xiang5Xiasheng Ding6Ni Zhang7School of Mechanical and Electrical Engineering, Lanzhou University of Technology, Lanzhou, ChinaProvincial Key Institute of Robotics, School of Automation and Electrical Engineering, Zhejiang University of Science and Technology, Hangzhou, ChinaProvincial Key Institute of Robotics, School of Automation and Electrical Engineering, Zhejiang University of Science and Technology, Hangzhou, ChinaProvincial Key Institute of Robotics, School of Automation and Electrical Engineering, Zhejiang University of Science and Technology, Hangzhou, ChinaSchool of Mechanical and Electrical Engineering, Lanzhou University of Technology, Lanzhou, ChinaProvincial Key Institute of Robotics, School of Automation and Electrical Engineering, Zhejiang University of Science and Technology, Hangzhou, ChinaZhejiang Chenlong Sawing Machine Company Ltd., Lishui, ChinaZhejiang Chenlong Sawing Machine Company Ltd., Lishui, ChinaComplete and high-quality labeled dataset is indispensable for image classification. Considering the often arduous task of data labeling, clustering algorithms are commonly utilized in the preliminary stages of data classification to perform preliminary categorization. Traditional clustering algorithms, such as K-means and Gaussian Mixture Model, often struggle to effectively cluster images. In this paper, we propose a novel image clustering method utilizing teacher-student model and autoencoder. Specifically, the teacher model is a fully connected autoencoder and the student model consists of multiple convolutional autoencoders. We firstly obtain sub-datasets by applying the K-means algorithm to cluster the output of the teacher model, then utilize the student model to achieve clustering through an iterative data exchange method where the same data can converge. In addition, the proposed method can filter low-quality data to a certain extent by recording the frequency of exchanges because clustering is achieved during multiple exchanges of data. Experiments on a synthetic dataset and three benchmark image datasets (MNIST, FashionMNIST and USPS) show that our method can achieve satisfactory clustering results.https://ieeexplore.ieee.org/document/10255740/Image clusteringteacher-student modelautoencoder |
| spellingShingle | Zhengshun Fei Haibo Gong Junhao Guo Jinglong Wang Wuyin Jin Xinjian Xiang Xiasheng Ding Ni Zhang Image Clustering: Utilizing Teacher-Student Model and Autoencoder IEEE Access Image clustering teacher-student model autoencoder |
| title | Image Clustering: Utilizing Teacher-Student Model and Autoencoder |
| title_full | Image Clustering: Utilizing Teacher-Student Model and Autoencoder |
| title_fullStr | Image Clustering: Utilizing Teacher-Student Model and Autoencoder |
| title_full_unstemmed | Image Clustering: Utilizing Teacher-Student Model and Autoencoder |
| title_short | Image Clustering: Utilizing Teacher-Student Model and Autoencoder |
| title_sort | image clustering utilizing teacher student model and autoencoder |
| topic | Image clustering teacher-student model autoencoder |
| url | https://ieeexplore.ieee.org/document/10255740/ |
| work_keys_str_mv | AT zhengshunfei imageclusteringutilizingteacherstudentmodelandautoencoder AT haibogong imageclusteringutilizingteacherstudentmodelandautoencoder AT junhaoguo imageclusteringutilizingteacherstudentmodelandautoencoder AT jinglongwang imageclusteringutilizingteacherstudentmodelandautoencoder AT wuyinjin imageclusteringutilizingteacherstudentmodelandautoencoder AT xinjianxiang imageclusteringutilizingteacherstudentmodelandautoencoder AT xiashengding imageclusteringutilizingteacherstudentmodelandautoencoder AT nizhang imageclusteringutilizingteacherstudentmodelandautoencoder |