The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
ASR(자동 음성 인식)과 MT(기계 번역) 시스템을 결합한 음성 번역 시스템은 화자의 유창성 및 인식 오류로 인한 중복되고 관련 없는 정보로 인해 성능이 저하됩니다. 본 논문에서는 ASR 오류와 유창성을 제거하고 의미 있는 문구를 추출하는 음성 통합을 통해 음성 인식 결과를 번역하는 새로운 접근 방식을 제안합니다. 통합 접근 방식은 ASR 1-best의 단어 추출을 통해 음성 요약에서 분리됩니다. 우리는 혼란 네트워크(CN)에 대한 통합 접근 방식을 확장하고 TED 연설을 사용하여 성능을 테스트했으며 통합 결과가 원래 ASR 결과와 비교하여 더 의미 있는 문구를 보존한다는 것을 확인했습니다. 우리는 통합 기술을 음성 번역에 적용했습니다. 통합 기반 음성 번역 성능을 테스트하기 위해 RT04의 중국어 방송 뉴스(BN) 음성을 인식하고 통합한 후 번역했습니다. 통합 기반 번역은 부분 번역이므로 통합을 통한 음성 번역 결과는 음성의 모든 단어가 번역되는 표준과 직접 비교할 수 없습니다. 병합을 통해 생성된 단어 네트워크에서 추출된 가장 유사한 단어 집합과 비교하여 부분 번역에 대한 새로운 평가 프레임워크를 제안하고자 합니다. 점진적인 요약 골드 스탠다드 번역. 통합 기반 MT 결과의 성능은 다음을 사용하여 평가되었습니다. 블루. 우리는 또한 제안 정보 보존 정확성 (IPAccy) 및 보존 정확도 의미 (MPAccy) 통합 및 통합 기반 MT를 평가합니다. 통합이 음성 번역 성능에 기여했음을 확인했습니다.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
부
Chiori HORI, Bing ZHAO, Stephan VOGEL, Alex WAIBEL, Hideki KASHIOKA, Satoshi NAKAMURA, "Consolidation-Based Speech Translation and Evaluation Approach" in IEICE TRANSACTIONS on Information,
vol. E92-D, no. 3, pp. 477-488, March 2009, doi: 10.1587/transinf.E92.D.477.
Abstract: The performance of speech translation systems combining automatic speech recognition (ASR) and machine translation (MT) systems is degraded by redundant and irrelevant information caused by speaker disfluency and recognition errors. This paper proposes a new approach to translating speech recognition results through speech consolidation, which removes ASR errors and disfluencies and extracts meaningful phrases. A consolidation approach is spun off from speech summarization by word extraction from ASR 1-best. We extended the consolidation approach for confusion network (CN) and tested the performance using TED speech and confirmed the consolidation results preserved more meaningful phrases in comparison with the original ASR results. We applied the consolidation technique to speech translation. To test the performance of consolidation-based speech translation, Chinese broadcast news (BN) speech in RT04 were recognized, consolidated and then translated. The speech translation results via consolidation cannot be directly compared with gold standards in which all words in speech are translated because consolidation-based translations are partial translations. We would like to propose a new evaluation framework for partial translation by comparing them with the most similar set of words extracted from a word network created by merging gradual summarizations of the gold standard translation. The performance of consolidation-based MT results was evaluated using BLEU. We also propose Information Preservation Accuracy (IPAccy) and Meaning Preservation Accuracy (MPAccy) to evaluate consolidation and consolidation-based MT. We confirmed that consolidation contributed to the performance of speech translation.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E92.D.477/_p
부
@ARTICLE{e92-d_3_477,
author={Chiori HORI, Bing ZHAO, Stephan VOGEL, Alex WAIBEL, Hideki KASHIOKA, Satoshi NAKAMURA, },
journal={IEICE TRANSACTIONS on Information},
title={Consolidation-Based Speech Translation and Evaluation Approach},
year={2009},
volume={E92-D},
number={3},
pages={477-488},
abstract={The performance of speech translation systems combining automatic speech recognition (ASR) and machine translation (MT) systems is degraded by redundant and irrelevant information caused by speaker disfluency and recognition errors. This paper proposes a new approach to translating speech recognition results through speech consolidation, which removes ASR errors and disfluencies and extracts meaningful phrases. A consolidation approach is spun off from speech summarization by word extraction from ASR 1-best. We extended the consolidation approach for confusion network (CN) and tested the performance using TED speech and confirmed the consolidation results preserved more meaningful phrases in comparison with the original ASR results. We applied the consolidation technique to speech translation. To test the performance of consolidation-based speech translation, Chinese broadcast news (BN) speech in RT04 were recognized, consolidated and then translated. The speech translation results via consolidation cannot be directly compared with gold standards in which all words in speech are translated because consolidation-based translations are partial translations. We would like to propose a new evaluation framework for partial translation by comparing them with the most similar set of words extracted from a word network created by merging gradual summarizations of the gold standard translation. The performance of consolidation-based MT results was evaluated using BLEU. We also propose Information Preservation Accuracy (IPAccy) and Meaning Preservation Accuracy (MPAccy) to evaluate consolidation and consolidation-based MT. We confirmed that consolidation contributed to the performance of speech translation.},
keywords={},
doi={10.1587/transinf.E92.D.477},
ISSN={1745-1361},
month={March},}
부
TY - JOUR
TI - Consolidation-Based Speech Translation and Evaluation Approach
T2 - IEICE TRANSACTIONS on Information
SP - 477
EP - 488
AU - Chiori HORI
AU - Bing ZHAO
AU - Stephan VOGEL
AU - Alex WAIBEL
AU - Hideki KASHIOKA
AU - Satoshi NAKAMURA
PY - 2009
DO - 10.1587/transinf.E92.D.477
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E92-D
IS - 3
JA - IEICE TRANSACTIONS on Information
Y1 - March 2009
AB - The performance of speech translation systems combining automatic speech recognition (ASR) and machine translation (MT) systems is degraded by redundant and irrelevant information caused by speaker disfluency and recognition errors. This paper proposes a new approach to translating speech recognition results through speech consolidation, which removes ASR errors and disfluencies and extracts meaningful phrases. A consolidation approach is spun off from speech summarization by word extraction from ASR 1-best. We extended the consolidation approach for confusion network (CN) and tested the performance using TED speech and confirmed the consolidation results preserved more meaningful phrases in comparison with the original ASR results. We applied the consolidation technique to speech translation. To test the performance of consolidation-based speech translation, Chinese broadcast news (BN) speech in RT04 were recognized, consolidated and then translated. The speech translation results via consolidation cannot be directly compared with gold standards in which all words in speech are translated because consolidation-based translations are partial translations. We would like to propose a new evaluation framework for partial translation by comparing them with the most similar set of words extracted from a word network created by merging gradual summarizations of the gold standard translation. The performance of consolidation-based MT results was evaluated using BLEU. We also propose Information Preservation Accuracy (IPAccy) and Meaning Preservation Accuracy (MPAccy) to evaluate consolidation and consolidation-based MT. We confirmed that consolidation contributed to the performance of speech translation.
ER -