The Second DIHARD Diarization Challenge: Dataset, task, and baselines

By Neville Ryant, Kenneth Ward Church, C. Cieri, Alejandrina Cristia, Jun Du, Sriram Ganapathy, Mark Y. Liberman. 2019

203 citations recorded by the source corpus.

This paper introduces the second DIHARD challenge, the second in a series of speaker diarization challenges intended to improve the robustness of diarization systems to variation in recording equipment, noise conditions, and conversational domain. The challenge comprises four tracks evaluating diarization performance under two input conditions (single channel vs. multi-channel) and two segmentation conditions (diarization from a reference speech segmentation vs. diarization from scratch). In order to prevent participants from overtuning to a particular combination of recording conditions and conversational domain, recordings are drawn from a variety of sources ranging from read audiobooks to meeting speech, to child language acquisition recordings, to dinner parties, to web video. We describe the task and metrics, challenge design, datasets, and baseline systems for speech enhancement, speech activity detection, and diarization.

Subfields: Artificial Intelligence, Evaluation & Benchmarks, Speech & Audio, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Evaluation & Benchmarks

In Speech & Audio

In Signal Processing

Citations in the corpus

CHiME-6 Challenge: Tackling Multispeaker Speech Recognition for Unsegmented Recordings (2020, positive): “Track 1 is a rerun of the CHiME-5 challenge [34] and Track 2 is similar to the “Diarization from multichannel audio using system SAD” track of the DIHARD II challenge [35], with the following key differences:”
Towards Neural Diarization for Unlimited Numbers of Speakers Using Global and Local Attractors (2021, background): “Even if it is adapted to datasets that contain more speakers than four ( e.g ., DIHARD [17, 3]), it is hard to produce diarization results of more than four speakers.”
A Review of Speaker Diarization: Recent Advances with Deep Learning (2021, methodology): “, it is part of the DIHARD and CHiME both the baseline and the top-performing systems [84, 85, 86, 89, 90].”
End-to-End Speaker Diarization for an Unknown Number of Speakers with Encoder-Decoder Based Attractors (2020, methodology): “For the DIHARD dataset, we also used the Jaccard error rate (JER), and we did not use collar tolerance, following the regulation of the second DIHARD challenge [32].”
Speaker recognition based on deep learning: An overview (2020, methodology): “real scenarios [308, 309]. ‹ DIHARD1,2: The DIHARD challenge intended to improve the robustness of diarization systems to variation in recording equipment, noise conditions, and conversational domain [201, 310]. ‹ AMI: The AMI meeting corpus is a multi-modal data set consisting of 100 hours of meeti”
Spot the conversation: speaker diarisation in the wild (2020, background): “The second DIHARD [6, 7] challenge provides a baseline system based on the JHU submission of the first DIHARD challenge.”
Encoder-Decoder Based Attractor Calculation for End-to-End Neural Diarization (2021, methodology): “Such models have been trained on the basis of clean speech (or time-frequency masks calculated from clean speech); thus, they cannot be trained on real mixtures like DIHARD datasets [33], [34].”