NSTL国家科技图书文献中心

1. TRAIN LONG AND TEST LONG: LEVERAGING FULL DOCUMENT CONTEXTS IN SPEECH PROCESSING NSTL国家科技图书文献中心

William Chen | Takatomo Kano ... - 《ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2024), Vol.17: Seoul, South Korea.14-19 April 2024》 - IEEE International Conference on Acoustics, Speech and Signal Processing - 2024, - 13066～13070 - 共5页

摘要：The quadratic memory complexity of self-attention has generally restricted Transformer-based models to utterance-based speech processing, preventing models from leveraging long-form contexts. A common...

关键词： Long-form ASR | Speech Summarization | Speech Translation | Self-supervised Learning

2. SPEECH SUMMARIZATION OF LONG SPOKEN DOCUMENT: IMPROVING MEMORY EFFICIENCY OF SPEECH/TEXT ENCODERS NSTL国家科技图书文献中心

Takatomo Kano | Atsunori Ogawa ... - 《2023 IEEE International Conference on Acoustics, Speech and Signal Processing: ICASSP 2023, Rhodes Island, Greece, 4-10 June 2023, [v.7]》 - IEEE International Conference on Acoustics, Speech and Signal Processing - 2023, - 5521～5525 - 共5页

摘要：Speech summarization requires processing several minute-long speech sequences to allow exploiting the whole context of a spoken document. A conventional approach is a cascade of automatic speech recog...

关键词： end-to-end modeling | long spoken document | memory efficient encoders | dual speech/text encoder

3. Summarize While Translating: Universal Model With Parallel Decoding for Summarization and Translation NSTL国家科技图书文献中心

Takatomo Kano | Atsunori Ogawa ... - 《2023 IEEE Automatic Speech Recognition and Understanding Workshop: IEEE Automatic Speech Recognition and Understanding Workshop (ASRU 2023), 16-20 Dec. 2023, Taipei, Taiwan》 - IEEE Automatic Speech Recognition and Understanding Workshop - 2023, - 1～8 - 共8页

摘要：Recently, multi-decoder and universal models have attracted increased interest in speech and language processing as they allow learning common representations across tasks. These models learn a common...

关键词： Training | Conferences | Performance gain | Multitasking | Decoding | Task analysis | Speech processing

4. Espnet-Summ: Introducing a Novel Large Dataset, Toolkit, and a Cross-Corpora Evaluation of Speech Summarization Systems NSTL国家科技图书文献中心

Roshan Sharma | William Chen ... - 《2023 IEEE Automatic Speech Recognition and Understanding Workshop: IEEE Automatic Speech Recognition and Understanding Workshop (ASRU 2023), 16-20 Dec. 2023, Taipei, Taiwan》 - IEEE Automatic Speech Recognition and Understanding Workshop - 2023, - 1～8 - 共8页

摘要：Speech summarization has garnered significant interest and progressed rapidly over the past few years. In particular, end-to-end models have recently emerged as a competitive alternative to cascade sy...

关键词： Training | Conferences | Oral communication | Interviews | Task analysis | Automatic speech recognition

5. Integrating Multiple ASR Systems into NLP Backend with Attention Fusion NSTL国家科技图书文献中心

Takatomo Kano | Atsunori Ogawa ... - 《2022 IEEE International Conference on Acoustics, Speech and Signal Processing: 47th IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Virtual Conference, 7-13 May 2022, 22-27 May 2022, Singapore》 - IEEE International Conference on Acoustics, Speech and Signal Processing - 2022, - 6237～6241 - 共5页

摘要：Spoken language processing (SLP) systems such as speech summarization and translation can be achieved by cascade models. It combines an automatic speech recognition (ASR) frontend and a natural langua...

关键词： Conferences | Signal processing | Natural language processing | Acoustics | Machine translation | Speech processing | Task analysis

NSTL主题词： back-end | FUSION

6. Integrating Multiple ASR Systems into NLP Backend with Attention Fusion NSTL国家科技图书文献中心

Takatomo Kano | Atsunori Ogawa ... - 《ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing, Vol.9: Virtual Conference, 7-13 May 2022 and Singapore, 22-27 May 2022》 - IEEE International Conference on Acoustics, Speech and Signal Processing - 2022, - 6237～6241 - 共5页

摘要：Spoken language processing (SLP) systems such as speech summarization and translation can be achieved by cascade models. It combines an automatic speech recognition (ASR) frontend and a natural langua...

关键词： Speech summarization | Speech translation | Automatic speech recognition | Attention fusion | ROVER

NSTL主题词： NINL gene | Natural Language Processing | Speech recognition | aquifer storage and recovery | back-end | FUSION | Fusion reactions | fusion (melting) | Fusion, biological

7. BLSTM-Based Confidence Estimation for End-to-End Speech Recognition NSTL国家科技图书文献中心

Atsunori Ogawa | Naohiro Tawara ... - 《2021 IEEE International Conference on Acoustics, Speech and Signal Processing: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 6-11 June 2021, Virtual Conference, Toronto, Ontario, Canada》 - IEEE International Conference on Acoustics, Speech and Signal Processing - 2021, - 6383～6387 - 共5页

摘要：Confidence estimation, in which we estimate the reliability of each recognized token (e.g., word, sub-word, and character) in automatic speech recognition (ASR) hypotheses and detect incorrectly recog...

关键词： Error analysis | Conferences | Estimation | Speech recognition | Signal processing | Decoding | Character recognition

NSTL主题词： end-to-end | confidence

8. Transformer-Based Direct Speech-To-Speech Translation with Transcoder NSTL国家科技图书文献中心

Takatomo Kano | Sakriani Sakti ... - 《2021 IEEE Spoken Language Technology Workshop: IEEE Spoken Language Technology Workshop (SLT), 19-22 Jan. 2021, Shenzhen, China》 - Spoken Language Technology Workshop - 2021, - 958～965 - 共8页

摘要：Traditional speech translation systems use a cascade manner that concatenates speech recognition (ASR), machine translation (MT), and text-to-speech (TTS) synthesis to translate speech from one langua...

关键词： Recurrent neural networks | Speech recognition | Syntactics | Predictive models | Machine translation | Task analysis | Spectrogram

NSTL主题词： transcoders | Mentoring

9. BLSTM-Based Confidence Estimation for End-to-End Speech Recognition NSTL国家科技图书文献中心

Atsunori Ogawa | Naohiro Tawara ... - 《ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing, Vol.10: Toronto, Ontario, Canada.6-11 June 2021》 - IEEE International Conference on Acoustics, Speech and Signal Processing - 2021, - 6383～6387 - 共5页

摘要：Confidence estimation, in which we estimate the reliability of each recognized token (e.g., word, sub-word, and character) in automatic speech recognition (ASR) hypotheses and detect incorrectly recog...

关键词： End-to-end (E2E) speech recognition | confidence estimation | imbalanced datasets | bidirectional long short-term memory (BLSTM) | auxiliary features

NSTL主题词： Speech recognition | end-to-end | confidence

10. Attention-Based Multi-Hypothesis Fusion for Speech Summarization NSTL国家科技图书文献中心

Takatomo Kano | Atsunori Ogawa ... - 《2021 IEEE Automatic Speech Recognition and Understanding Workshop: IEEE Automatic Speech Recognition and Understanding Workshop (ASRU), 13-17 Dec. 2021, Cartagena, Colombia》 - IEEE Automatic Speech Recognition and Understanding Workshop - 2021, - 487～494 - 共8页

摘要：Speech summarization, which generates a text summary from speech, can be achieved by combining automatic speech recognition (ASR) and text summarization (TS). With this cascade approach, we can exploi...

摘要：The quadratic memory complexity of self-attention has generally restricted Transformer-based models to utterance-based speech processing, preventing models from leveraging long-form contexts. A common...

关键词： Long-form ASR | Speech Summarization | Speech Translation | Self-supervised Learning

摘要：Speech summarization requires processing several minute-long speech sequences to allow exploiting the whole context of a spoken document. A conventional approach is a cascade of automatic speech recog...

关键词： end-to-end modeling | long spoken document | memory efficient encoders | dual speech/text encoder

摘要：Recently, multi-decoder and universal models have attracted increased interest in speech and language processing as they allow learning common representations across tasks. These models learn a common...

关键词： Training | Conferences | Performance gain | Multitasking | Decoding | Task analysis | Speech processing

摘要：Speech summarization has garnered significant interest and progressed rapidly over the past few years. In particular, end-to-end models have recently emerged as a competitive alternative to cascade sy...

关键词： Training | Conferences | Oral communication | Interviews | Task analysis | Automatic speech recognition

摘要：Spoken language processing (SLP) systems such as speech summarization and translation can be achieved by cascade models. It combines an automatic speech recognition (ASR) frontend and a natural langua...

关键词： Conferences | Signal processing | Natural language processing | Acoustics | Machine translation | Speech processing | Task analysis

NSTL主题词： back-end | FUSION

摘要：Spoken language processing (SLP) systems such as speech summarization and translation can be achieved by cascade models. It combines an automatic speech recognition (ASR) frontend and a natural langua...

关键词： Speech summarization | Speech translation | Automatic speech recognition | Attention fusion | ROVER

NSTL主题词： NINL gene | Natural Language Processing | Speech recognition | aquifer storage and recovery | back-end | FUSION | Fusion reactions | fusion (melting) | Fusion, biological

摘要：Confidence estimation, in which we estimate the reliability of each recognized token (e.g., word, sub-word, and character) in automatic speech recognition (ASR) hypotheses and detect incorrectly recog...

关键词： Error analysis | Conferences | Estimation | Speech recognition | Signal processing | Decoding | Character recognition

NSTL主题词： end-to-end | confidence

摘要：Traditional speech translation systems use a cascade manner that concatenates speech recognition (ASR), machine translation (MT), and text-to-speech (TTS) synthesis to translate speech from one langua...

关键词： Recurrent neural networks | Speech recognition | Syntactics | Predictive models | Machine translation | Task analysis | Spectrogram

NSTL主题词： transcoders | Mentoring

摘要：Confidence estimation, in which we estimate the reliability of each recognized token (e.g., word, sub-word, and character) in automatic speech recognition (ASR) hypotheses and detect incorrectly recog...

关键词： End-to-end (E2E) speech recognition | confidence estimation | imbalanced datasets | bidirectional long short-term memory (BLSTM) | auxiliary features

NSTL主题词： Speech recognition | end-to-end | confidence

摘要：Speech summarization, which generates a text summary from speech, can be achieved by combining automatic speech recognition (ASR) and text summarization (TS). With this cascade approach, we can exploi...

关键词： Training | Conferences | Bit error rate | Transformers | Task analysis | Automatic speech recognition

NSTL主题词： Speech | FUSION

4008-161-200 800-990-8900

国家科技图书文献中心