Reverberant speech recognition based on denoising autoencoder

By Takaaki Ishii, Hiroki Komiyama, Takahiro Shinozaki, Yasuo Horiuchi, Shingo Kuroiwa. 2013

105 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Representation & Self-Supervised Learning, 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 Computer Vision and Pattern Recognition

In Representation & Self-Supervised Learning

In Speech & Audio

In Signal Processing

Citations in the corpus

SoundStream: An End-to-End Neural Audio Codec (2021, methodology): “Audio enhancement – Deep neural networks have been applied to different audio enhancement tasks, ranging from denoising [17]–[21] to dereverberation [22], [23], lossy coding denoising [2] and frequency bandwidth extension [1], [24].”
A practical tutorial on autoencoders for nonlinear feature fusion: Taxonomy, models, software and guidelines (2018, methodology): “A similar procedure is followed in [143], although in this case the problem present in the speech samples is reverberation.”
Deep Learning for Environmentally Robust Speech Recognition (2017, methodology): “types databases front (mappingbased) mag [84] deep CNN add. TIMIT log mag [70] DBM add.&con. 2nd CHiME log power [69] DBM add. TIMIT [71] DBM add. Aurora-2 mel [68] DSAE add. Japan. speech logMel [73] DBM con. CENSREC-4 [80] LSTM-DRNN add.&con. 2nd CHiME [42] LSTM-DRNN con. REVERB MFCC [74] ”
The MERL/MELCO/TUM system for the REVERB Challenge using Deep Recurrent Neural Network Feature Enhancement (2014, methodology): “Similar to previous studies on DAE in ASR [5, 19, 20], we directly use the output of the auto-encoder as features, not the hidden layer activations – this allows usage in a ‘black-box’ scenario where only the ASR feature extraction but not the ASR back-end is known.”
Deep recurrent de-noising auto-encoder and blind de-reverberation for reverberated speech recognition (2014, methodology): “Our first goal is to show that spectral enhancement by a de-noising auto-encoder (DAE) [2,11] generalizes to real reverberated speech in unseen acoustic conditions.”
Toward Robust Neural Machine Translation for Noisy Input Sequences (2017, methodology): “Data augmentation has also been used speciﬁcally to improve robustness to noisy inputs as in our work, such as research on speech recognition under noisy conditions [20] and translating spelling mistakes [5, 6].”
Deep neural network based spectral feature mapping for robust speech recognition (2015, methodology): “Most previous work focused on learning clean version of MFCC using either denoising autoencoders [12, 13] or RNNs [14].”
Speech feature denoising and dereverberation via deep autoencoders for noisy reverberant speech recognition (2014, background): “Index Terms— robust speech recognition, feature denoising, denoising autoencoder, deep neural network”
Reverberant speech recognition combining deep neural networks and deep autoencoders augmented with a phone-class feature (2015, methodology): “Recognition of reverberant speech is performed combining “standard” DNN-HMM [14] decoding and a feature enhancement through deep autoencoder (DAE) [9, 10, 15].”
A deep auto-encoder based low-dimensional feature extraction from FFT spectral envelopes for statistical parametric speech synthesis (2016, methodology): “Deep auto-encoder based bottleneck features have been used by several groups for ASR [15, 16] and a deep denoising autoencoder has also verified for noise-robust ASR [17] or reverberant ASR tasks [18, 19].”