Open AccessPosted Content
Improving Phoneme segmentation with Recurrent Neural Networks.
TLDR
This work proposes a novel unsupervised algorithm based on sequence prediction models such as Markov chains and recurrent neural network that tries to learn the dynamics of speech in the MFCC space and hypothesize boundaries from local maxima in the prediction error.Abstract:
Phonemic segmentation of speech is a critical step of speech recognition systems. We propose a novel unsupervised algorithm based on sequence prediction models such as Markov chains and recurrent neural network. Our approach consists in analyzing the error profile of a model trained to predict speech features frame-by-frame. Specifically, we try to learn the dynamics of speech in the MFCC space and hypothesize boundaries from local maxima in the prediction error. We evaluate our system on the TIMIT dataset, with improvements over similar methods.read more
Citations
More filters
Proceedings ArticleDOI
Self-Supervised Contrastive Learning for Unsupervised Phoneme Segmentation.
TL;DR: In this article, a self-supervised representation learning model is proposed for unsupervised phoneme boundary detection, which is optimized to identify spectral changes in the signal using the Noise-Contrastive Estimation principle.
Posted Content
Gate Activation Signal Analysis for Gated Recurrent Neural Networks and Its Correlation with Phoneme Boundaries
TL;DR: The temporal structure of gate activation signals inside the gated recurrent neural networks is highly correlated with the phoneme boundaries, and this correlation is further verified by a set of experiments for phoneme segmentation.
Posted Content
Towards unsupervised phone and word segmentation using self-supervised vector-quantized neural networks
TL;DR: This work constrain pretrained self-supervised vector-quantized (VQ) neural networks so that blocks of contiguous feature vectors are assigned to the same code, thereby giving a variable-rate segmentation of the speech into discrete units.
Posted Content
Phoneme Boundary Detection using Learnable Segmental Features.
TL;DR: The authors proposed a neural architecture coupled with a parameterized structured loss function to learn segmental representations for the task of phoneme boundary detection, which achieved state-of-the-art performance in terms of F1 and R-value.
Proceedings ArticleDOI
Attacking the problem of continuous speech segmentation into basic units
TL;DR: The paper considers the algorithm of continuous speech segmentation into basic units, namely phonemes, certain combination of phoneme and pauses, based on speech signal transformation into a two-dimensional image, i.e. an autocorrelation portrait.
References
More filters
Journal ArticleDOI
Long short-term memory
TL;DR: A novel, efficient, gradient based method called long short-term memory (LSTM) is introduced, which can learn to bridge minimal time lags in excess of 1000 discrete-time steps by enforcing constant error flow through constant error carousels within special units.
Journal Article
Dropout: a simple way to prevent neural networks from overfitting
TL;DR: It is shown that dropout improves the performance of neural networks on supervised learning tasks in vision, speech recognition, document classification and computational biology, obtaining state-of-the-art results on many benchmark data sets.
Journal ArticleDOI
Finding Structure in Time
TL;DR: A proposal along these lines first described by Jordan (1986) which involves the use of recurrent links in order to provide networks with a dynamic memory and suggests a method for representing lexical categories and the type/token distinction is developed.
Proceedings ArticleDOI
k-means++: the advantages of careful seeding
TL;DR: By augmenting k-means with a very simple, randomized seeding technique, this work obtains an algorithm that is Θ(logk)-competitive with the optimal clustering.
Journal ArticleDOI
Comparison of parametric representations for monosyllabic word recognition in continuously spoken sentences
S. Davis,Paul Mermelstein +1 more
TL;DR: In this article, several parametric representations of the acoustic signal were compared with regard to word recognition performance in a syllable-oriented continuous speech recognition system, and the emphasis was on the ability to retain phonetically significant acoustic information in the face of syntactic and duration variations.