Convolutional Neural Networks for Distant Speech Recognition

By Pawel Swietojanski, Arnab Ghoshal, S. Renals. 2014

244 citations recorded by the source corpus.

Subfields: Artificial Intelligence, 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 Signal Processing

Citations in the corpus

Recent advances in convolutional neural networks (2015, neutral): “308], since they are well suited to exploit the correlations in both time and frequency domains through the local connectivity and are capable of capturing frequency shift in human speech signals. In [306, 309, 310], they achieve lower speech recognition errors by applying CNN on Mel lter bank featu”
Residual LSTM: Design of a Deep Recurrent Architecture for Distant Speech Recognition (2017, background): “For example, convolutional neural networks (CNN) which has been huge success in image classification and detection were effective to reduce environmental and speaker variability in acoustic features [6, 7, 8, 9, 10].”
Highway long short-term memory RNNS for distant speech recognition (2015, neutral): “Since we do not exclude the overlapping speech segments during model training, in addition to results on the full eval set, we also show results on a subset that only contains the non-overlapping speech segments as [14].”
Analysis of CNN-based speech recognition system using raw speech as input (2015, methodology): “In most of the recent convolutional neural networks-based systems proposed in the literature, the input features are either conventional cepstral-based features [5] or spectral-based representations, such as Mel filterbank coefficients [10].”
Improved MVDR Beamforming Using Single-Channel Mask Prediction Networks (2016, background): “Recent work on multi-microphone databases such as the AMI database [1, 2, 3] and recent challenges such as REVERB [4], CHiME-2 [5], and CHiME-3 [6] indicate that performing beamforming of multiple microphone data improves performance as compared to other direct techniques for speech recognition.”
Deep Convolutional Neural Networks with Layer-Wise Context Expansion and Attention (2016, methodology): “Various CNN structures [6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24] have been evaluated for speech recognition.”
Very Deep Convolutional Neural Networks for Noise Robust Speech Recognition (2016, methodology): “The effect of CNNs was first verified in image classification [11], [12], [15] and then applied to speech recognition [9], [11], [13], [14], [16], [17].”
Perception, Guidance, and Navigation for Indoor Autonomous Drone Racing Using Deep Learning (2018, background): “In this letter, we propose to combine a CNN-based deeplearning and the LOS guidance for the IROS-style indoor drone racing [1], which only allows onboard computing.”
Neural network based spectral mask estimation for acoustic beamforming (2016, background): “And despite recent attempts to take advantage of multi-channel data within a (convolutional) DNN or even training a network directly on multi-channel waveforms, model-based beamforming still proved to be superior [1] [2].”
A deep learning approach to detection of splicing and copy-move forgeries in images (2016, background): “In recent years, deep neural networks, such as Deep Belief Network [9], Deep Auto Encoder [10] and Convolutional Neural Network (CNN) [11], have shown to be capable of extracting complex statistical dependencies from high-dimensional sensory inputs and efficiently learning their hierarchical represe”
Speech acoustic modeling from raw multichannel waveforms (2015, methodology): “In the multichannel case (Table 2) we compare the waveform CNN against stacked mel-tb features similar to [13, 14] (but without pooling across channels) and against mel-tb features computed on the output of a delay-and-sum beamformer.”
Convolutional Neural Networks-based continuous speech recognition using raw speech signal (2015, methodology): “More specifically, use of context-dependent phonemes [9, 10]; use of spectral features as opposed to cepstral features [4, 11]; CNN-based system with 4295978-1-4673-6997-8/15/$31.00 ©2015 IEEE ICASSP 2015 mel filter bank energies as input [12, 13]; combination of different features [14]; CNN-based p”
Deep Learning for Environmentally Robust Speech Recognition (2017, unknown): “eamformers and post-ﬁlters for speech enhancement, joint front- and back-end multi-channel ASR systems have recently attracted considerable attention with a goal of decreasing the WER directly [131], [132], [134]. In [130], the individual features extracted from each microphone channel are concatena”
New Era for Robust Speech Recognition (2017, background): “It permits the substitution of the solution algorithm from loopy belief propagation [52] if desired.”
Automatic Sleep Stage Scoring with Single-Channel EEG Using Convolutional Neural Networks (2016, methodology): “Again, there are characteristic examples from speech and acoustics in [6, 19, 27, 20, 10].”
Deep learning approach for active classification of electrocardiogram signals (2016, background): “Typical deep learning architectures include deep belief networks (DBNs) [26], stacked autoencoder (SAE) [54], convolutional neural networks (CNNs) [52], .”