Obstructive sleep apnea (OSA) is a common disorder characterized by recurring breathing pauses during caused blockage of the upper airway (UA). OSA generally diagnosed through costly procedure requiring an overnight stay patient at hospital. This has led to proposing less procedures based on analysis patients’ facial images and voice recordings help in detection severity assessment. In this paper we investigate use both image speech processing estimate apnea-hypopnea index, AHI (which...

The rapid increase in web services and mobile apps, which collect personal data from users, has also increased the risk that their privacy may be severely compromised. In particular, increasing variety of spoken language interfaces voice assistants empowered by vertiginous breakthroughs deep learning have prompted important concerns European Union terms preserving speech data. For instance, an attacker can record users impersonate them to obtain access systems require identification. By...

Sleep apnea (OSA) is a common sleep disorder characterized by recurring breathing pauses during caused blockage of the upper airway (UA). The altered UA structure or function in OSA speakers has led to hypothesize automatic analysis speech for assessment. In this paper we critically review several approaches using and machine learning techniques detection, discuss limitations that can arise when diagnostic applications. A large database including 426 male Spanish suspected suffer derived...

Obstructive Sleep Apnea (OSA) is a sleep breathing disorder affecting at least 3-7% of male adults and 2-5% female between 30 70 years. It causes recurrent partial or total obstruction episodes the level pharynx which cessation breath during sleep. The number per hour, known as Apnea-Hypopnea Index (AHI), along with degree daytime sleepiness, determine severity OSA. Usually, OSA diagnosed Unit in hospital by time-consuming polysomnography (PSG) test. Based on expected impact anatomical...

This paper describes an exploratory technique to identify mild dementia by assessing the degree of speech deficits. A total twenty participants were used for this experiment, ten patients with a diagnosis and like healthy control. The audio session each subject was recorded following methodology developed present study. Prosodic features in elderly controls measured using automatic prosodic analysis on reading task. novel method carried out gather twelve over samples. best classification...

Background: Obstructive sleep apnea (OSA) is a common disorder characterized by frequent cessation of breathing lasting 10 seconds or longer. The diagnosis OSA performed through an expensive procedure, which requires overnight stay at the hospital. This has led to several proposals based on analysis patients' facial images and speech recordings as attempt develop simpler cheaper methods diagnose OSA. Objective: objective this study was analyze possible relationships between features female...

This paper describes a comparison between hybrid and end-to-end Automatic Speech Recognition (ASR) systems, which were evaluated on the IberSpeech-RTVE 2020 Speech-to-Text Transcription Challenge. Deep Neural Networks (DNNs) are becoming most promising technology for ASR at present. In last few years, traditional models have been compared to other systems in terms of accuracy efficiency. We contribute two different approaches: system based DNN-HMM state-of-the-art Lattice-Free Maximum Mutual...

The fast increase of web services and mobile apps, which collect personal data from users, increases the risk that their privacy may be severely compromised. In particular, increasing variety spoken language interfaces voice assistants empowered by vertiginous breakthroughs in Deep Learning are prompting important concerns European Union to preserve speech privacy. For instance, an attacker can record users impersonate them get access systems requiring identification. Hacking speaker...

Inspired by successful work in forensic speaker identification, this presents a higher level system for text-independent recognition means of the temporal trajectories formant frequencies linguistic units. Feature extraction from unit-dependent provides very flexible able to be applied different scenarios. At fine-grained level, it is possible provide calibrated likelihood ratio per unit under analysis (extremely useful applications such as forensics), and at coarse-grained individual...

<sec> <title>BACKGROUND</title> Obstructive sleep apnea (OSA) is a common disorder characterized by frequent cessation of breathing lasting 10 seconds or longer. The diagnosis OSA performed through an expensive procedure, which requires overnight stay at the hospital. This has led to several proposals based on analysis patients’ facial images and speech recordings as attempt develop simpler cheaper methods diagnose OSA. </sec> <title>OBJECTIVE</title> objective this study was analyze...