DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | 유창동 | - |
dc.contributor.author | Yoon, Eunseop | - |
dc.contributor.author | 윤은섭 | - |
dc.date.accessioned | 2024-07-25T19:31:14Z | - |
dc.date.available | 2024-07-25T19:31:14Z | - |
dc.date.issued | 2023 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=1045904&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/320674 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 전기및전자공학부, 2023.8,[v, 35 p. :] | - |
dc.description.abstract | Automatic Speech Recognition (ASR) is a task that converts a spoken language into written text, and these systems have attained unprecedented performance with large speech models pre-trained based on self-supervised speech representation learning. However, these pre-trained speech models suffer from representational bias as they tend to better represent those prominent accents (i.e., native (L1) English accent) in the pre-training speech corpus than less represented accents, resulting in a deteriorated performance for non-native (L2) English accents. Although there have been some approaches to mitigate this issue, all of these methods require updating the pre-trained model weights. In this paper, we propose Information Theoretic Adversarial Prompt Tuning (INTapt), which introduces prompts concatenated to the original input that can re-modulate the attention of the pre-trained model such that the corresponding input resembles a native (L1) English speech without updating the backbone weights. INTapt is trained simultaneously in the following two manners: (1) adversarial training to reduce accent feature dependence between the original input and the prompt-concatenated input and (2) training to minimize CTC loss for improving ASR performance to a prompt-concatenated input. Experimental results show that INTapt improves the performance of L2 English and increases feature similarity between L2 and L1 accents. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | 음성 인식▼a프롬프트 튜닝▼a도메인 적 | - |
dc.subject | Automatic speech recognition▼aPrompt tuning▼aDomain adaptation | - |
dc.title | Deep learning based approach for enhanced non-native speech recognition | - |
dc.title.alternative | 비원어민의 음성인식 향상을 위한 딥러닝을 활용한 접근법 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :전기및전자공학부, | - |
dc.contributor.alternativeauthor | Yoo, Changdong | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.