Real-life voice activity detection with LSTM Recurrent Neural Networks and an application to Hollywood movies

By Florian Eyben, Felix Weninger, Stefano Squartini, Björn Schuller. 2013

236 citations recorded by the source corpus.

Subfields: Computational Mechanics, 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 Computational Mechanics

1. J. Urgen Schmidhuber
2. Maja Pantic
3. Björn Schuller
4. Yoshua Bengio
5. Michel François Valstar
6. Timur R. Almaev
7. Dougal Maclaurin
8. Hugo Larochelle

In Speech & Audio

1. Yoshua Bengio
2. Mirco Ravanelli
3. Philemon Brakel
4. Maurizio Omologo
5. J. Urgen Schmidhuber
6. Maja Pantic
7. Björn Schuller
8. Florian Eyben

In Signal Processing

1. Yoshua Bengio
2. Mirco Ravanelli
3. Björn Schuller
4. Maja Pantic
5. Philemon Brakel
6. Florian Eyben
7. Maurizio Omologo
8. Michel François Valstar

Citations in the corpus

• Light Gated Recurrent Units for Speech Recognition (2018, background): “Several works have already highlighted the effectiveness of RNNs in various speech processing tasks, such as speech recognition [35]–[39], speech enhancement [40], speech separation [41], [42] as well as speech activity detection [43].”
• Deep learning in neural networks: An overview (2014, background): “, 2014), voice activity detection (Eyben et al., 2013), optical character recognition (Breuel et al.”
• AVEC 2014: 3D Dimensional Affect and Depression Recognition Challenge (2014, methodology): “A set of baseline features is provided for three different versions of segmentation: First, a voice activity detector [7] was applied to obtain a segmentation based on”
• A review of modularization techniques in artificial neural networks (2019, background): “LSTM has shown state-of-the-art results in sequence modeling and different real-world problems (Stollenga et al. 2015; Eyben et al. 2013; Soutner and Müller 2013).”
• Neural Dialogue Context Online End-of-Turn Detection (2018, methodology): “We divided each data set into speech units and non-speech units using an LSTM-RNN based SAD (Eyben et al., 2013) 1 where #calls, #utterances, and #turns represent number of calls, utterances and end-of-turn points, respectively.”
• Improving Speech Recognition by Revising Gated Recurrent Units (2017, background): “Several works have already highlighted the effectiveness of RNNs in various speech processing tasks such as speech recognition [11, 12, 13, 14, 15], speech enhancement [16], speech separation [17, 18] as well as speech activity detection [19].”
• AVEC 2013: the continuous audio/visual emotion and depression recognition challenge (2013, methodology): “sal features from automatically detected voice activity [6] segments worked better.”