The DIRHA-ENGLISH corpus and related tasks for distant-speech recognition in domestic environments

By Mirco Ravanelli, L. Cristoforetti, Roberto Gretter, Marco Pellin, Alessandro Sosi, Maurizio Omologo. 2015

61 citations recorded by the source corpus.

This paper introduces the contents and the possible usage of the DIRHA-ENGLISH multi-microphone corpus, recently realized under the EC DIRHA project. The reference scenario is a domestic environment equipped with a large number of microphones and microphone arrays distributed in space. The corpus is composed of both real and simulated material, and it includes 12 US and 12 UK English native speakers. Each speaker uttered different sets of phonetically-rich sentences, newspaper articles, conversational speech, keywords, and commands. From this material, a large set of 1-minute sequences was generated, which also includes typical domestic background noise as well as inter/intra-room reverberation effects. Dev and test sets were derived, which represent a very precious material for different studies on multi-microphone speech processing and distant-speech recognition. Various tasks and corresponding Kaldi recipes have already been developed. The paper reports a first set of baseline results obtained using different techniques, including Deep Neural Networks (DNN), aligned with the state-of-the-art at international level.

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, neutral): “The CHiME-1 [15] and CHiME-2 [16] challenges and corpora have contributed to popularizing research on this topic, together with the DICIT [17], Sweet-Home [18], and DIRHA [19] corpora.”
Multi-Task Self-Supervised Learning for Robust Speech Recognition (2020, positive): “Additive noises are extracted from FreeSound and DIRHA datasets [19, 20] and include both background and non-stationary noises such as alarms, door knocks, telephone ringing, television, and many others.”
Light Gated Recurrent Units for Speech Recognition (2018, methodology): “To validate our model in a more realistic scenario, a set of experiments was also conducted in distant-talking conditions with the DIRHA-English corpus1 [54].”
Learning Problem-agnostic Speech Representations from Multiple Self-supervised Tasks (2019, methodology): “Next, to assess our approach in more challenging noisy and reverberant conditions, in Section 4.3 we use the DIRHA dataset [36].”
A network of deep neural networks for Distant Speech Recognition (2017, background): “More details on this corpus and on the impulse responses adopted in this work can be found in [33,34].”
Interpretable Convolutional Filters with SincNet (2018, methodology): “3 reports the speech recognition performance obtained by CNN and SincNet using the TIMIT and the DIRHA dataset [41].”
Improving Speech Recognition by Revising Gated Recurrent Units (2017, background): “More details on this dataset and on the impulse responses can be found in [36, 37].”
VoiceHome-2, an extended corpus for multichannel speech processing in real homes (2019, unknown): “The DIRHA-English corpus (Ravanelli et al., 2015) and the Sweet-Home corpus (Vacher et al.”
A French Corpus for Distant-Microphone Speech Processing in Real Homes (2016, unknown): “1The authors of [26] plan to distribute it via the LDC for a fee.”
Self-Supervised Speech Representation Learning: A Review (2022, background): “Studies leverage Hub5, DIRHA, and CHiME-5 to measure the efficacy of representations in ASR.”
The Pytorch-kaldi Speech Recognition Toolkit (2018, methodology): “To validate our model in a more challenging scenario, experiments were also conducted in distant-talking conditions with the DIRHA-English dataset(4) [36,37].”
Batch-normalized joint training for DNN-based distant speech recognition (2016, unknown): “For more details see [48, 47].”
An analysis of environment, microphone and data simulation mismatches in robust speech recognition (2017, unknown): “…progress thanks to real speech corpora collected for various application scenarios such as voice command for cars (Hansen et al., 2001), smart homes (Ravanelli et al., 2015), or tablets (Barker et al., 2015), and automatic transcription of lectures (Lamel et al., 1994), meetings (Renals et al.,…”
The fifth 'CHiME' Speech Separation and Recognition Challenge: Dataset, task and baselines (2018, unknown): “The CHiME-1 [14] and CHiME-2 [15] challenges and corpora have contributed to popularizing research on this topic, together with the DICIT [16], Sweet-Home [17], and DIRHA [18] corpora.”
Twin Regularization for online speech recognition (2018, methodology): “To validate our model in a more challenging scenario, experiments were also conducted in distant-talking conditions with the DIRHA-English dataset [32].”
Learning Speaker Representations with Mutual Information (2018, methodology): “In this case, all the clean signals were convoluted with a different impulse response [35], that was sampled from the DIRHA dataset [36, 37].”
Realistic Multi-Microphone Data Simulation for Distant Speech Recognition (2016, methodology): “Other simulated corpora have been released under the CHIL project [27] and, more recently, under the EU DIRHA project [28, 29, 30, 31, 32].”