Multiple flow‐based knowledge transfer via adversarial networks
The authors propose a new knowledge transfer method coupled with a generative adversarial network (GAN) when multiple‐flow‐based knowledge is considered in a teacher–student framework using a residual network (ResNet). In this method, several independent discriminators adapting multilayer‐perceptron...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2019-09-01
|
| Series: | Electronics Letters |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/el.2019.1874 |
| _version_ | 1828287266327363584 |
|---|---|
| author | D. Yeo J.‐H. Bae |
| author_facet | D. Yeo J.‐H. Bae |
| author_sort | D. Yeo |
| collection | DOAJ |
| description | The authors propose a new knowledge transfer method coupled with a generative adversarial network (GAN) when multiple‐flow‐based knowledge is considered in a teacher–student framework using a residual network (ResNet). In this method, several independent discriminators adapting multilayer‐perceptron‐based structures were designed for flow‐based knowledge transfer. The proposed GAN‐based optimisation alternatively updates the multiple discriminators and a student ResNet such that the flow‐based features of the student ResNet are generated as closely as possible to the real features of a teacher ResNet. The experiments demonstrate that the student ResNet trained using the proposed method more accurately captures the distribution of the flow‐based teacher knowledge than the l2‐distance‐based training method. In addition, the proposed method provided better classification accuracy than the existing GAN‐based knowledge transfer method. |
| first_indexed | 2024-04-13T09:45:45Z |
| format | Article |
| id | doaj.art-7a515dac5ba9495d83fb954ff943f244 |
| institution | Directory Open Access Journal |
| issn | 0013-5194 1350-911X |
| language | English |
| last_indexed | 2024-04-13T09:45:45Z |
| publishDate | 2019-09-01 |
| publisher | Wiley |
| record_format | Article |
| series | Electronics Letters |
| spelling | doaj.art-7a515dac5ba9495d83fb954ff943f2442022-12-22T02:51:46ZengWileyElectronics Letters0013-51941350-911X2019-09-01551898999210.1049/el.2019.1874Multiple flow‐based knowledge transfer via adversarial networksD. Yeo0J.‐H. Bae1Electronics and Telecommunications Research InstituteDaejeonRepublic of KoreaElectronics and Telecommunications Research InstituteDaejeonRepublic of KoreaThe authors propose a new knowledge transfer method coupled with a generative adversarial network (GAN) when multiple‐flow‐based knowledge is considered in a teacher–student framework using a residual network (ResNet). In this method, several independent discriminators adapting multilayer‐perceptron‐based structures were designed for flow‐based knowledge transfer. The proposed GAN‐based optimisation alternatively updates the multiple discriminators and a student ResNet such that the flow‐based features of the student ResNet are generated as closely as possible to the real features of a teacher ResNet. The experiments demonstrate that the student ResNet trained using the proposed method more accurately captures the distribution of the flow‐based teacher knowledge than the l2‐distance‐based training method. In addition, the proposed method provided better classification accuracy than the existing GAN‐based knowledge transfer method.https://doi.org/10.1049/el.2019.1874multiple flow‐based knowledge transfergenerative adversarial networkteacher–student frameworkresidual networkindependent discriminatorsmultilayer‐perceptron‐based structures |
| spellingShingle | D. Yeo J.‐H. Bae Multiple flow‐based knowledge transfer via adversarial networks Electronics Letters multiple flow‐based knowledge transfer generative adversarial network teacher–student framework residual network independent discriminators multilayer‐perceptron‐based structures |
| title | Multiple flow‐based knowledge transfer via adversarial networks |
| title_full | Multiple flow‐based knowledge transfer via adversarial networks |
| title_fullStr | Multiple flow‐based knowledge transfer via adversarial networks |
| title_full_unstemmed | Multiple flow‐based knowledge transfer via adversarial networks |
| title_short | Multiple flow‐based knowledge transfer via adversarial networks |
| title_sort | multiple flow based knowledge transfer via adversarial networks |
| topic | multiple flow‐based knowledge transfer generative adversarial network teacher–student framework residual network independent discriminators multilayer‐perceptron‐based structures |
| url | https://doi.org/10.1049/el.2019.1874 |
| work_keys_str_mv | AT dyeo multipleflowbasedknowledgetransferviaadversarialnetworks AT jhbae multipleflowbasedknowledgetransferviaadversarialnetworks |