Skip to Main content Skip to Navigation
Conference papers

What Makes a Speaker Recognizable in TV Broadcast? Going Beyond Speaker Identification Error Rate

Abstract : Speaker identification approaches for TV broadcast are usually evaluated and compared based on global error rates derived from the overall duration of missed detection, false alarm and confusion. Based on the analysis of the output of the systems submitted to the final round of the French evaluation campaign REPERE, this paper highlights the fact that these average met-rics lead to the incorrect intuition that current state-of-the-art algorithms partially recognize all speakers. Setting aside incorrect diarization and adverse acoustic conditions, we show that their performance is in fact essentially bi-modal: in a given show, either all speech turns of a speaker are correctly identified or none of them are. We then proceed with trying to understand and explain this behavior, through perfomance prediction experiments. These experiments show that the most discriminant speaker characteristics are – first – their total speech duration in the current show and – then only – the amount of training data available to build their acoustic model.
Complete list of metadatas

Cited literature [27 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01433205
Contributor : Sylvain Meignier <>
Submitted on : Thursday, April 6, 2017 - 8:59:44 AM
Last modification on : Monday, February 10, 2020 - 6:14:06 PM
Document(s) archivé(s) le : Friday, July 7, 2017 - 12:30:14 PM

File

Charlet2015.pdf
Publisher files allowed on an open archive

Identifiers

  • HAL Id : hal-01433205, version 1

Citation

Delphine Charlet, Johann Poignant, Hervé Bredin, Corinne Fredouille, Sylvain Meignier. What Makes a Speaker Recognizable in TV Broadcast? Going Beyond Speaker Identification Error Rate. ERRARE Workshop, a satellite event of Interspeech 2015., 2015, Sinaia, Romania. ⟨hal-01433205⟩

Share

Metrics

Record views

268

Files downloads

66