DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | 신진우 | - |
dc.contributor.author | Kim, Minkyu | - |
dc.contributor.author | 김민규 | - |
dc.date.accessioned | 2024-07-25T19:30:43Z | - |
dc.date.available | 2024-07-25T19:30:43Z | - |
dc.date.issued | 2023 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=1045714&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/320526 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 김재철AI대학원, 2023.8,[iii, 18 p. :] | - |
dc.description.abstract | existing supervised losses (such as cross-entropy) restrain the intra-class variation and limit the capability of learning rich representations. This issue becomes more severe when pre-training datasets are class-imbalanced or coarse-labeled. To address the problem, we propose a new representation learning method, named Renyi supervised contrastive learning (RenyiSCL), which can effectively learn transferable representation using a labeled dataset. Our main idea is to use the recently proposed self-supervised Renyi contrastive learning in the supervised setup. We show that RenyiSCL can mitigate the class-collapse problem by contrasting features with both instance-wise and class-wise information. Through experiments on the ImageNet dataset, we show that RenyiSCL outperforms all supervised and self-supervised methods under various transfer learning tasks. In particular, we also validate the effectiveness of RenyiSCL under class-imbalanced or coarse-labeled datasets. | - |
dc.description.abstract | A mighty goal of representation learning is to train a feature that can transfer to various tasks or datasets. A conventional approach is to pre-train a neural network on a large-scale labeled dataset, e.g., ImageNet, and use its feature for downstream tasks. However, the feature often lacks transferability due to the class-collapse issue | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | 지도 학습▼a표현식 학습▼a대조 학습▼a전이 학습 | - |
dc.subject | Supervised learning▼aRepresentation learning▼aContrastive learning▼aTransfer learning | - |
dc.title | Renyi supervised contrastive learning for transferable representation | - |
dc.title.alternative | 전이 가능한 표현식을 위한 레니 지도 대조 학습 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :김재철AI대학원, | - |
dc.contributor.alternativeauthor | Shin, Jinwoo | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.