Publication result detail

BERTraffic: BERT-based Joint Speaker Role and Speaker Change Detection for Air Traffic Control Communications

ZULUAGA-GOMEZ, J.; SARFJOO, S.; PRASAD, A.; NIGMATULINA, I.; MOTLÍČEK, P.; ONDŘEJ, K.; OHNEISER, O.; HELMKE, H.

Original Title

BERTraffic: BERT-based Joint Speaker Role and Speaker Change Detection for Air Traffic Control Communications

English Title

BERTraffic: BERT-based Joint Speaker Role and Speaker Change Detection for Air Traffic Control Communications

Type

Paper in proceedings (conference paper)

Original Abstract

Automatic speech recognition (ASR) allows transcribing the communications between air traffic controllers (ATCOs) and aircraft pilots. The transcriptions are used later to extract ATC named entities, e.g., aircraft callsigns. One common challenge is speech activity detection (SAD) and speaker diarization (SD). In the failure condition, two or more segments remain in the same recording, jeopardizing the overall performance. We propose a system that combines SAD and a BERT model to perform speaker change detection and speaker role detection (SRD) by chunking ASR transcripts, i.e., SD with a defined number of speakers together with SRD. The proposed model is evaluated on real-life public ATC databases. Our BERT SD model baseline reaches up to 10% and 20% token-based Jaccard error rate (JER) in public and private ATC databases. We also achieved relative improvements of 32% and 7.7% in JERs and SD error rate (DER), respectively, compared to VBx, a well-known SD system.1

English abstract

Automatic speech recognition (ASR) allows transcribing the communications between air traffic controllers (ATCOs) and aircraft pilots. The transcriptions are used later to extract ATC named entities, e.g., aircraft callsigns. One common challenge is speech activity detection (SAD) and speaker diarization (SD). In the failure condition, two or more segments remain in the same recording, jeopardizing the overall performance. We propose a system that combines SAD and a BERT model to perform speaker change detection and speaker role detection (SRD) by chunking ASR transcripts, i.e., SD with a defined number of speakers together with SRD. The proposed model is evaluated on real-life public ATC databases. Our BERT SD model baseline reaches up to 10% and 20% token-based Jaccard error rate (JER) in public and private ATC databases. We also achieved relative improvements of 32% and 7.7% in JERs and SD error rate (DER), respectively, compared to VBx, a well-known SD system.1

Keywords

Text-based speaker diarization, speaker change detection, speaker role detection, air traffic control communications, chunking

Key words in English

Text-based speaker diarization, speaker change detection, speaker role detection, air traffic control communications, chunking

Authors

ZULUAGA-GOMEZ, J.; SARFJOO, S.; PRASAD, A.; NIGMATULINA, I.; MOTLÍČEK, P.; ONDŘEJ, K.; OHNEISER, O.; HELMKE, H.

RIV year

2024

Released

09.01.2023

Publisher

IEEE Signal Processing Society

Location

Doha

ISBN

978-1-6654-7189-3

Book

IEEE Spoken Language Technology Workshop, SLT 2022 - Proceedings

Pages from

633

Pages to

640

Pages count

8

URL

https://ieeexplore.ieee.org/document/10022718

BibTex

@inproceedings{BUT185192,
  author="ZULUAGA-GOMEZ, J. and SARFJOO, S. and PRASAD, A. and NIGMATULINA, I. and MOTLÍČEK, P. and ONDŘEJ, K. and OHNEISER, O. and HELMKE, H.",
  title="BERTraffic: BERT-based Joint Speaker Role and Speaker Change Detection for Air Traffic Control Communications",
  booktitle="IEEE Spoken Language Technology Workshop, SLT 2022 - Proceedings",
  year="2023",
  pages="633--640",
  publisher="IEEE Signal Processing Society",
  address="Doha",
  doi="10.1109/SLT54892.2023.10022718",
  isbn="978-1-6654-7189-3",
  url="https://ieeexplore.ieee.org/document/10022718"
}

Documents

zuluaga-gomez_SLT2023_2023_10022718