Learning hidden unit contributions for unsupervised acoustic model adaptation
This work presents a broad study on the adaptation of neural network acoustic models by
means of learning hidden unit contributions (LHUC)-a method that linearly re-combines …
means of learning hidden unit contributions (LHUC)-a method that linearly re-combines …
Detection of acoustic impulse events in voice applications using a neural network
SPV Ebenezer - US Patent 10,475,471, 2019 - Google Patents
In accordance with embodiments of the present disclosure, an integrated circuit for
implementing at least a portion of an audio device may include an audio output configured …
implementing at least a portion of an audio device may include an audio output configured …
Learning to adapt: a meta-learning approach for speaker adaptation
The performance of automatic speech recognition systems can be improved by adapting an
acoustic model to compensate for the mismatch between training and testing conditions, for …
acoustic model to compensate for the mismatch between training and testing conditions, for …
Structured output layer with auxiliary targets for context-dependent acoustic modelling
In previous work we have introduced a multi-task training technique for neural network
acoustic modelling, in which context-dependent and context-independent targets are jointly …
acoustic modelling, in which context-dependent and context-independent targets are jointly …
SAT-LHUC: Speaker adaptive training for learning hidden unit contributions
P Swietojanski, S Renais - 2016 IEEE International Conference …, 2016 - ieeexplore.ieee.org
This paper extends learning hidden unit contributions (LHUC) unsupervised speaker
adaptation with speaker adaptive training (SAT). Contrary to other SAT approaches, the …
adaptation with speaker adaptive training (SAT). Contrary to other SAT approaches, the …
Differentiable pooling for unsupervised acoustic model adaptation
P Swietojanski, S Renals - IEEE/ACM Transactions on Audio …, 2016 - ieeexplore.ieee.org
We present a deep neural network (DNN) acoustic model that includes parametrised and
differentiable pooling operators. Unsupervised acoustic model adaptation is cast as the …
differentiable pooling operators. Unsupervised acoustic model adaptation is cast as the …
Detection of acoustic impulse events in voice applications
SPV Ebenezer - US Patent 10,242,696, 2019 - Google Patents
In accordance with embodiments of the present disclosure, an integrated circuit for
implementing at least a portion of an audio device may include an audio output configured …
implementing at least a portion of an audio device may include an audio output configured …
Combination of multiple acoustic models with unsupervised adaptation for lecture speech transcription
Automatic speech recognition systems (ASR) have achieved considerable progress in real
applications because of skilled design of the architecture with advanced techniques and …
applications because of skilled design of the architecture with advanced techniques and …
Learning representations for speech recognition using artificial neural networks
P Swietojanski - 2016 - era.ed.ac.uk
Learning representations is a central challenge in machine learning. For speech
recognition, we are interested in learning robust representations that are stable across …
recognition, we are interested in learning robust representations that are stable across …
Learning to adapt: meta-learning approaches for speaker adaptation
O Klejch - 2020 - era.ed.ac.uk
The performance of automatic speech recognition systems degrades rapidly when there is a
mismatch between training and testing conditions. One way to compensate for this mismatch …
mismatch between training and testing conditions. One way to compensate for this mismatch …