DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Ro, Yong Man | - |
dc.contributor.advisor | 노용만 | - |
dc.contributor.author | Kim, Eun Sung | - |
dc.date.accessioned | 2021-05-13T19:33:07Z | - |
dc.date.available | 2021-05-13T19:33:07Z | - |
dc.date.issued | 2020 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=911320&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/284711 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 전기및전자공학부, 2020.2,[iii, 22 p. :] | - |
dc.description.abstract | With the recent rapid developments of deep learning, deep neural networks have been applied to various computer vision tasks such as object classification, detection, and segmentation, and have shown superior performance. Deep neural networks, however, suffers from the problem called "catastrophic forgetting" that is a tendency of forgetting previously learned tasks when deep neural networks learn new tasks. Catastrophic forgetting is considered as a fundamental limitation of deep neural networks. Knowledge distillation-based methods have been proposed to alleviate catastrophic forgetting problem, but the performance is still not good. In this paper, we propose a Task-selective Autoencoder (TsAE) to improve the performance of class incremental learning of the conventional knowledge distillation-based method. The proposed Task-selective Autoencoder has a simple architecture and a small number of trainable parameters. The experimental results show that the application of Task-selective autoencoder to the existing knowledge distillation-based methods effectively improves classification performance. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Deep learning▼adeep neural network▼acatastrophic forgetting▼aincremental learning▼aclass incremental learning▼aclassification▼aautoencoder | - |
dc.subject | 딥러닝▼a딥 뉴럴 네트워크▼a파괴적 망각▼a점진 학습▼a클래스 점진 학습▼a분류▼a오토인코터 | - |
dc.title | Class incremental learning with task-selective autoencoder | - |
dc.title.alternative | 태스크 선택 오토인코더를 통한 클래스 점진 학습 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :전기및전자공학부, | - |
dc.contributor.alternativeauthor | 김은성 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.