Improving Speaker Assignment in Speaker-Attributed ASR for Real Meeting Applications
Not on the map. Why?
Abstract (arXiv)
Past studies on end-to-end meeting transcription have focused on model architecture and have mostly been evaluated on simulated meeting data. We present a novel study aiming to optimize the use of a Speaker-Attributed ASR (SA-ASR) system in real-life scenarios, such as the AMI meeting corpus, for improved speaker assignment of speech segments. First, we propose a pipeline tailored to real-life applications involving Voice Activity Detection (VAD), Speaker Diarization (SD), and SA-ASR. Second, we advocate using VAD output segments to fine-tune the SA-ASR model, considering that it is also applied to VAD segments during test, and show that this results in a relative reduction of Speaker Error Rate (SER) up to 28%. Finally, we explore strategies to enhance the extraction of the speaker embedding templates used as inputs by the SA-ASR system. We show that extracting them from SD output rather than annotated speaker segments results in a relative SER reduction up to 20%.
Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 26 of its 42 references that are in the map. First posted on arXiv 11 Mar 2024.
Builds on
26 of 42 references.
References · 26
- 2018SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processingcited by 1,989
- 2020ECAPA-TDNN: Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verificationcited by 628
- 2021A Review of Speaker Diarization: Recent Advances with Deep Learningcited by 124
- 2020Serialized Output Training for End-to-End Overlapped Speech Recognitioncited by 87
- 2021M2MeT: The ICASSP 2022 Multi-Channel Multi-Party Meeting Transcription Challengecited by 71
- 2020Target-Speaker Voice Activity Detection: a Novel Approach for Multi-Speaker Diarization in a Dinner Party Scenariocited by 68
- 2018gpuRIR: A Python Library for Room Impulse Response Simulation with GPU Accelerationcited by 46
- 2020End-to-End Multi-speaker Speech Recognition with Transformercited by 42
- 2020Joint Speaker Counting, Speech Recognition, and Speaker Identification for Overlapped Speech of Any Number of Speakerscited by 41
- 2021End-to-End Speaker-Attributed ASR with Transformercited by 35
- 2021Large-Scale Pre-Training of End-to-End Multi-Talker ASR for Meeting Transcription with Single Distant Microphonecited by 33
- 2018A Purely End-to-end System for Multi-speaker Speech Recognitioncited by 30
14 more
- 2020Investigation of End-To-End Speaker-Attributed ASR for Continuous Multi-Talker Recordingscited by 28
- 2020VoxSRC 2020: The Second VoxCeleb Speaker Recognition Challengecited by 27
- 2019Advances in Online Audio-Visual Meeting Transcriptioncited by 23
- 2021Transcribe-to-Diarize: Neural Speaker Diarization for Unlimited Number of Speakers using End-to-End Speaker-Attributed ASRcited by 22
- 2023The CHiME-7 DASR Challenge: Distant Meeting Transcription with Multiple Devices in Diverse Scenarioscited by 22
- 2020Microsoft Speaker Diarization System for the VoxCeleb Speaker Recognition Challenge 2020cited by 18
- 2019VoxSRC 2019: The first VoxCeleb Speaker Recognition Challengecited by 14
- 2022Streaming Speaker-Attributed ASR with Token-Level Speaker Embeddingscited by 14
- 2021Multi-Speaker ASR Combining Non-Autoregressive Conformer CTC and Conditional Speaker Chaincited by 12
- 2022Simulating realistic speech overlaps improves multi-talker ASRcited by 10
- 2021Hypothesis Stitcher for End-to-End Speaker-attributed ASR on Long-form Multi-talker Recordingscited by 7
- 2022A Comparative Study on Speaker-attributed Automatic Speech Recognition in Multi-party Meetingscited by 5
- 2020Fast CRDNN: Towards on Site Training of Mobile Construction Machinescited by 4
- 2023End-to-end Multichannel Speaker-Attributed ASR: Speaker Guided Decoder and Input Feature Analysiscited by 3
Built on by
Citing papers per year: 2024: 1; 2025: 1; 2026: 1.
3 papers.