arXiv subjects

Jordi Luque

Publications and source records attributed to Jordi Luque.

35 records · Page 2Linked to original sources

arXiv2021

Voice Quality and Pitch Features in Transformer-Based Speech Recognition

Jitter and shimmer measurements have shown to be carriers of voice quality and prosodic information which enhance the performance of tasks like speaker recognition, diarization or automatic speech recognition (ASR). However, such features have been seldom used in the context of neural-based ASR, where spectral features often prevail. In this work, we study the effects of incorporating voice quality and pitch features altogether and separately to a Transformer-based ASR model, with the intuition that the attention mechanisms might exploit latent prosodic traits. For doing so, we propose separated convolutional front-ends for prosodic and spectral features, showing that this architectural choice yields better results than simple concatenation of such pitch and voice quality features to mel-spectrogram filterbanks. Furthermore, we find mean Word Error Rate relative reductions of up to 5.6% with the LibriSpeech benchmark. Such findings motivate further research on the application of prosody knowledge for increasing the robustness of Transformer-based ASR.

cs.CL↗

arXiv2021

Influence of ASR and Language Model on Alzheimer's Disease Detection

Alzheimer's Disease is the most common form of dementia. Automatic detection from speech could help to identify symptoms at early stages, so that preventive actions can be carried out. This research is a contribution to the ADReSSo Challenge, we analyze the usage of a SotA ASR system to transcribe participant's spoken descriptions from a picture. We analyse the loss of performance regarding the use of human transcriptions (measured using transcriptions from the 2020 ADReSS Challenge). Furthermore, we study the influence of a language model -- which tends to correct non-standard sequences of words -- with the lack of language model to decode the hypothesis from the ASR. This aims at studying the language bias and get more meaningful transcriptions based only on the acoustic information from patients. The proposed system combines acoustic -- based on prosody and voice quality -- and lexical features based on the first occurrence of the most common words. The reported results show the effect of using automatic transcripts with or without language model. The best fully automatic system achieves up to 76.06 % of accuracy (without language model), significantly higher, 3 % above, than a system employing word transcriptions decoded using general purpose language models.

cs.CL↗

arXiv2021

Efficient Keyword Spotting by capturing long-range interactions with Temporal Lambda Networks

Models based on attention mechanisms have shown unprecedented speech recognition performance. However, they are computationally expensive and unnecessarily complex for keyword spotting, a task targeted to small-footprint devices. This work explores the application of Lambda networks, an alternative framework for capturing long-range interactions without attention, for the keyword spotting task. We propose a novel \textit{ResNet}-based model by swapping the residual blocks by temporal Lambda layers. Furthermore, the proposed architecture is built upon uni-dimensional temporal convolutions that further reduce its complexity. The presented model does not only reach state-of-the-art accuracies on the Google Speech Commands dataset, but it is 85% and 65% lighter than its Transformer-based (KWT) and convolutional (Res15) counterparts while being up to 100 times faster. To the best of our knowledge, this is the first attempt to explore the Lambda framework within the speech domain and therefore, we unravel further research of new interfaces based on this architecture.

eess.AS↗

arXiv2021

English Accent Accuracy Analysis in a State-of-the-Art Automatic Speech Recognition System

Nowadays, research in speech technologies has gotten a lot out thanks to recently created public domain corpora that contain thousands of recording hours. These large amounts of data are very helpful for training the new complex models based on deep learning technologies. However, the lack of dialectal diversity in a corpus is known to cause performance biases in speech systems, mainly for underrepresented dialects. In this work, we propose to evaluate a state-of-the-art automatic speech recognition (ASR) deep learning-based model, using unseen data from a corpus with a wide variety of labeled English accents from different countries around the world. The model has been trained with 44.5K hours of English speech from an open access corpus called Multilingual LibriSpeech, showing remarkable results in popular benchmarks. We test the accuracy of such ASR against samples extracted from another public corpus that is continuously growing, the Common Voice dataset. Then, we present graphically the accuracy in terms of Word Error Rate of each of the different English included accents, showing that there is indeed an accuracy bias in terms of accentual variety, favoring the accents most prevalent in the training corpus.

eess.AS↗

arXiv2021

BCN2BRNO: ASR System Fusion for Albayzin 2020 Speech to Text Challenge

This paper describes joint effort of BUT and Telefónica Research on development of Automatic Speech Recognition systems for Albayzin 2020 Challenge. We compare approaches based on either hybrid or end-to-end models. In hybrid modelling, we explore the impact of SpecAugment layer on performance. For end-to-end modelling, we used a convolutional neural network with gated linear units (GLUs). The performance of such model is also evaluated with an additional n-gram language model to improve word error rates. We further inspect source separation methods to extract speech from noisy environment (i.e. TV shows). More precisely, we assess the effect of using a neural-based music separator named Demucs. A fusion of our best systems achieved 23.33% WER in official Albayzin 2020 evaluations. Aside from techniques used in our final submitted systems, we also describe our efforts in retrieving high quality transcripts for training.

eess.AS↗

arXiv2021

Speech Enhancement for Wake-Up-Word detection in Voice Assistants

Keyword spotting and in particular Wake-Up-Word (WUW) detection is a very important task for voice assistants. A very common issue of voice assistants is that they get easily activated by background noise like music, TV or background speech that accidentally triggers the device. In this paper, we propose a Speech Enhancement (SE) model adapted to the task of WUW detection that aims at increasing the recognition rate and reducing the false alarms in the presence of these types of noises. The SE model is a fully-convolutional denoising auto-encoder at waveform level and is trained using a log-Mel Spectrogram and waveform reconstruction losses together with the BCE loss of a simple WUW classification network. A new database has been purposely prepared for the task of recognizing the WUW in challenging conditions containing negative samples that are very phonetically similar to the keyword. The database is extended with public databases and an exhaustive data augmentation to simulate different noises and environments. The results obtained by concatenating the SE with a simple and state-of-the-art WUW detectors show that the SE does not have a negative impact on the recognition rate in quiet environments while increasing the performance in the presence of noise, especially when the SE and WUW detector are trained jointly end-to-end.

eess.AS↗

arXiv2020

Convolutional Speech Recognition with Pitch and Voice Quality Features

The effects of adding pitch and voice quality features such as jitter and shimmer to a state-of-the-art CNN model for Automatic Speech Recognition are studied in this work. Pitch features have been previously used for improving classical HMM and DNN baselines, while jitter and shimmer parameters have proven to be useful for tasks like speaker or emotion recognition. Up to our knowledge, this is the first work combining such pitch and voice quality features with modern convolutional architectures, showing improvements up to 7% and 3% relative WER points, for the publicly available Spanish Common Voice and LibriSpeech 100h datasets, respectively. Particularly, our work combines these features with mel-frequency spectral coefficients (MFSCs) to train a convolutional architecture with Gated Linear Units (Conv GLUs). Such models have shown to yield small word error rates, while being very suitable for parallel processing for online streaming recognition use cases. We have added pitch and voice quality functionality to Facebook's wav2letter speech recognition framework, and we provide with such code and recipes to the community, to carry on with further experiments. Besides, to the best of our knowledge, our Spanish Common Voice recipe is the first public Spanish recipe for wav2letter.

eess.AS↗

arXiv2020

Transcription-Enriched Joint Embeddings for Spoken Descriptions of Images and Videos

In this work, we propose an effective approach for training unique embedding representations by combining three simultaneous modalities: image and spoken and textual narratives. The proposed methodology departs from a baseline system that spawns a embedding space trained with only spoken narratives and image cues. Our experiments on the EPIC-Kitchen and Places Audio Caption datasets show that introducing the human-generated textual transcriptions of the spoken narratives helps to the training procedure yielding to get better embedding representations. The triad speech, image and words allows for a better estimate of the point embedding and show an improving of the performance within tasks like image and speech retrieval, even when text third modality, text, is not present in the task.

cs.CV↗

arXiv2020

Input complexity and out-of-distribution detection with likelihood-based generative models

Likelihood-based generative models are a promising resource to detect out-of-distribution (OOD) inputs which could compromise the robustness or reliability of a machine learning system. However, likelihoods derived from such models have been shown to be problematic for detecting certain types of inputs that significantly differ from training data. In this paper, we pose that this problem is due to the excessive influence that input complexity has in generative models' likelihoods. We report a set of experiments supporting this hypothesis, and use an estimate of input complexity to derive an efficient and parameter-free OOD score, which can be seen as a likelihood-ratio, akin to Bayesian model comparison. We find such score to perform comparably to, or even better than, existing OOD detection approaches under a wide range of data sets, models, model sizes, and complexity estimates.

cs.LG↗

arXiv2019

Detection of speech events and speaker characteristics through photo-plethysmographic signal neural processing

The use of photoplethysmogram signal (PPG) for heart and sleep monitoring is commonly found nowadays in smartphones and wrist wearables. Besides common usages, it has been proposed and reported that person information can be extracted from PPG for other uses, like biometry tasks. In this work, we explore several end-to-end convolutional neural network architectures for detection of human's characteristics such as gender or person identity. In addition, we evaluate whether speech/non-speech events may be inferred from PPG signal, where speech might translate in fluctuations into the pulse signal. The obtained results are promising and clearly show the potential of fully end-to-end topologies for automatic extraction of meaningful biomarkers, even from a noisy signal sampled by a low-cost PPG sensor. The AUCs for best architectures put forward PPG wave as biological discriminant, reaching $79\%$ and $89.0\%$, respectively for gender and person verification tasks. Furthermore, speech detection experiments reporting AUCs around $69\%$ encourage us for further exploration about the feasibility of PPG for speech processing tasks.

eess.AS↗

arXiv2019

Seeing and Hearing Egocentric Actions: How Much Can We Learn?

Our interaction with the world is an inherently multimodal experience. However, the understanding of human-to-object interactions has historically been addressed focusing on a single modality. In particular, a limited number of works have considered to integrate the visual and audio modalities for this purpose. In this work, we propose a multimodal approach for egocentric action recognition in a kitchen environment that relies on audio and visual information. Our model combines a sparse temporal sampling strategy with a late fusion of audio, spatial, and temporal streams. Experimental results on the EPIC-Kitchens dataset show that multimodal integration leads to better performance than unimodal approaches. In particular, we achieved a 5.18% improvement over the state of the art on verb classification.

cs.CV↗

arXiv2019

How Much Does Audio Matter to Recognize Egocentric Object Interactions?

Sounds are an important source of information on our daily interactions with objects. For instance, a significant amount of people can discern the temperature of water that it is being poured just by using the sense of hearing. However, only a few works have explored the use of audio for the classification of object interactions in conjunction with vision or as single modality. In this preliminary work, we propose an audio model for egocentric action recognition and explore its usefulness on the parts of the problem (noun, verb, and action classification). Our model achieves a competitive result in terms of verb classification (34.26% accuracy) on a standard benchmark with respect to vision-based state of the art systems, using a comparatively lighter architecture.

cs.CV↗

arXiv2016

Emergence of linguistic laws in human voice

Linguistic laws constitute one of the quantitative cornerstones of modern cognitive sciences and have been routinely investigated in written corpora, or in the equivalent transcription of oral corpora. This means that inferences of statistical patterns of language in acoustics are biased by the arbitrary, language-dependent segmentation of the signal, and virtually precludes the possibility of making comparative studies between human voice and other animal communication systems. Here we bridge this gap by proposing a method that allows to measure such patterns in acoustic signals of arbitrary origin, without needs to have access to the language corpus underneath. The method has been applied to six different human languages, recovering successfully some well-known laws of human communication at timescales even below the phoneme and finding yet another link between complexity and criticality in a biological system. These methods further pave the way for new comparative studies in animal communication or the analysis of signals of unknown code.

physics.soc-ph↗

arXiv2014

Speech earthquakes: scaling and universality in human voice

Speech is a distinctive complex feature of human capabilities. In order to understand the physics underlying speech production, in this work we empirically analyse the statistics of large human speech datasets ranging several languages. We first show that during speech the energy is unevenly released and power-law distributed, reporting a universal robust Gutenberg-Richter-like law in speech. We further show that such earthquakes in speech show temporal correlations, as the interevent statistics are again power-law distributed. Since this feature takes place in the intra-phoneme range, we conjecture that the responsible for this complex phenomenon is not cognitive, but it resides on the physiological speech production mechanism. Moreover, we show that these waiting time distributions are scale invariant under a renormalisation group transformation, suggesting that the process of speech generation is indeed operating close to a critical point. These results are put in contrast with current paradigms in speech processing, which point towards low dimensional deterministic chaos as the origin of nonlinear traits in speech fluctuations. As these latter fluctuations are indeed the aspects that humanize synthetic speech, these findings may have an impact in future speech synthesis technologies. Results are robust and independent of the communication language or the number of speakers, pointing towards an universal pattern and yet another hint of complexity in human speech.

physics.soc-ph↗

arXiv2010

Horizontal visibility graphs: exact results for random time series

The visibility algorithm has been recently introduced as a mapping between time series and complex networks. This procedure allows to apply methods of complex network theory for characterizing time series. In this work we present the horizontal visibility algorithm, a geometrically simpler and analytically solvable version of our former algorithm, focusing on the mapping of random series (series of independent identically distributed random variables). After presenting some properties of the algorithm, we present exact results on the topological properties of graphs associated to random series, namely the degree distribution, clustering coefficient, and mean path length. We show that the horizontal visibility algorithm stands as a simple method to discriminate randomness in time series, since any random series maps to a graph with an exponential degree distribution of the shape P(k) = (1/3)(2/3)**(k-2), independently of the probability distribution from which the series was generated. Accordingly, visibility graphs with other P(k) are related to non-random series. Numerical simulations confirm the accuracy of the theorems for finite series. In a second part, we show that the method is able to distinguish chaotic series from i.i.d. theory, studying the following situations: (i) noise-free low-dimensional chaotic series, (ii) low-dimensional noisy chaotic series, even in the presence of large amounts of noise, and (iii) high-dimensional chaotic series (coupled map lattice), without needs for additional techniques such as surrogate data or noise reduction methods. Finally, heuristic arguments are given to explain the topological properties of chaotic series and several sequences which are conjectured to be random are analyzed.

physics.data-an↗

arXiv2009

The Visibility Graph: a new method for estimating the Hurst exponent of fractional Brownian motion

Fractional Brownian motion (fBm) has been used as a theoretical framework to study real time series appearing in diverse scientific fields. Because its intrinsic non-stationarity and long range dependence, its characterization via the Hurst parameter H requires sophisticated techniques that often yield ambiguous results. In this work we show that fBm series map into a scale free visibility graph whose degree distribution is a function of H. Concretely, it is shown that the exponent of the power law degree distribution depends linearly on H. This also applies to fractional Gaussian noises (fGn) and generic f^(-b) noises. Taking advantage of these facts, we propose a brand new methodology to quantify long range dependence in these series. Its reliability is confirmed with extensive numerical simulations and analytical developments. Finally, we illustrate this method quantifying the persistent behavior of human gait dynamics.

physics.data-an↗

arXiv2008

From time series to complex networks: the visibility graph

In this work we present a simple and fast computational method, the visibility algorithm, that converts a time series into a graph. The constructed graph inherits several properties of the series in its structure. Thereby, periodic series convert into regular graphs, and random series do so into random graphs. Moreover, fractal series convert into scale-free networks, enhancing the fact that power law degree distributions are related to fractality, something highly discussed recently. Some remarkable examples and analytical tools are outlined in order to test the method's reliability. Many different measures, recently developed in the complex network theory, could by means of this new approach characterize time series from a new point of view.

physics.data-an↗