The performance of traditional biometric identification systems is, as yet, unsatisfactory in certain applications. For this reason, other physiological or behavioral characteristics have recently been considered, using new electrical physical signals linked to a person's vital signs. This paper examines the phonocardiogram (PCG) from cardiac auscultation. idea is that PCG specific individual can be taken into consideration sign used system. More specifically, proposes preliminary study...

The recognition of emotional states is a relatively new technique in the field machine learning. paper presents study and performance results system for emotion classification using architecture distributed speech (DSR). features used were extracted by front-end ETSI Aurora eXtended mobile terminal compliance with ES 202-211 V1.1.1 standard. On basis time trend these parameters, over 3800 statistical parameters to characterize semantic units varying length (sentences words). Using WEKA...

The paper proposes a performance evaluation and comparison of G.729, AMR, fuzzy voice activity detection (FVAD) algorithms. was made using objective, psychoacoustic, subjective parameters. A highly varied speech database also set up to evaluate the extent which VADs depend on language, signal-to-noise ratio (SNR), or power level.

We study the problem of data gathering in wireless sensor networks and compare several approaches belonging to different research fields; particular, signal processing, compressive sensing, information theory, networking related techniques are investigated. Specifically, we derived a simple analytical model able predict energy efficiency reliability techniques. Moreover, carry out simulations validate our effectiveness above schemes by systematically sampling parameter space (i.e., number...

Sleep disorders are steadily increasing in the population and can significantly affect daily life. Low-cost noninvasive systems that assist diagnostic process will become increasingly widespread coming years. This work aims to investigate compare performance of machine learning-based classifiers for identification obstructive sleep apnea–hypopnea (OSAH) events, including apnea/non-apnea status classification, index (AHI) prediction, AHI severity classification. The dataset considered...

In this paper we propose a new lossless compression algorithm suitable for Internet of Things (IoT). The proposed algorithm, named RAKE, is based only on elementary counting operations and has low memory requirements, therefore it can be easily implemented in low-cost low-speed micro-controllers as those used IoT devices. Despite its simplicity, simulation results show that, the case sparse sequences, outperforms well-known algorithms such rar, gzip bzip2. Moreover, real-world data, RAKE...

Communication systems based on man-machine interaction play an increasingly important role in everyday life. Whereas these make life easier for many people, the hearing impaired they are indispensable. Automatic speech recognition particular, thanks to simultaneous display of text, reduce gap separating deaf, or general, from world hearing. The paper presents automatic aid system while behind wheel a car. It proposes algorithm, typical techniques, identification and signaling emergency...

In WSNs low complexity techniques for reducing the amount of bits flowing throughout network are needed to decrease bandwidth waste and increase lifetime. To this purpose, spatial correlation due high density deployed devices can be exploited. paper we propose a novel in-network aggregation technique based on Chinese Remainder Theorem (CRT), which exploits well known advantages source coding while using distributed approach with no need coordination among nodes. Also, an appropriate...

Nowadays, the Intracranial Pressure (ICP) monitoring has become most common method of investigation for both traumatic and chronic neural pathologies. ICP signals are typically triphasic, that is, in a single waveform, three subpeaks can be identified. This work outlines new algorithm to identify from recordings extract number 20 meaningful parameter trends. The validity implemented been proved through comparison between automatic identification by manually marked neurosurgeon. marking...

One of the challenging topics and design constraints in Wireless Sensor Networks (WSNs) is reduction energy consumption because, most application scenarios, replacement power resources sensor devices might be unfeasible. In order to minimize consumption, some nodes can put sleep during idle times wake up only when needed. Although it seems best way limit energy, other performance parameters such as network reliability have considered. a recent paper, we introduced new forwarding algorithm...

This paper focuses on a novel lossless compression algorithm which can be efficiently used for of electrocardiogram (ECG) signals. The proposed has low memory requirements and relies simple efficient encoding scheme implemented with elementary counting operations. Thus it easily even in resource constrained microcontrollers as those commonly several low-cost ECG monitoring systems. Despite its simplicity, simulation results carried out real-world signals show that the achieves higher ratios...

Spontaneous intracranial hypotension (SIH) is considered to be a very rare disease. It characterised by an orthostatic headache in the absence of past history trauma or dural puncture. SIH caused spontaneous spinal cerebrospinal fluid (CSF) leakage demonstrated neuroradiological studies most patients. Conservative treatment usually includes bed rest, hydration and administration caffeine steroids. However, when patient refractory conservative treatment, epidural blood patch (EBP) performed....

Spectrum sensing is an essential function of cognitive radio technology that can enable the reuse available resources by so-called secondary users without creating harmful interference with licensed users. The application machine learning techniques to spectrum has attracted considerable interest in literature. In this contribution, we study cooperative a network where multiple cooperate detect primary user. We introduce schemes based on deep neural network, which incorporate one-dimensional...

In this paper a new method for accurate frequency estimation of single sinusoid is described. The main feature the proposed method, mainly based on simple digital non-linear filter, its intrinsic low complexity development and use. in fact, needs about 2N <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">s</sub> arithmetic operations, where N number samples, has very small memory requirements even when high accuracy needed. Furthermore, we show...

In this paper, radio spectrum management is discussed in a multi-hop distributed network architecture scenario, dedicated for Internet of Things (IoT) applications, that aims to transform everyday objects smart objects. sense, important features Cognitive Radio Ad Hoc Networks (CRAHNs) are presented. The main interest given sensing functionality. Thus, the Multiple Signal Classification (MUSIC) technique and Minimum Variance Distortionless Response (MVDR) implemented. Obtained result...

Quality of service (QoS) is a crucial requirement in distributed applications. Internet Things architectures have become widely used approach many application domains, from Industry 4.0 to smart agriculture; thus, it develop appropriate methodologies for managing QoS such contexts. In an overcrowded spectrum scenario, cognitive radio technology could be effective methodology improving requirements. order evaluate the context network, we propose Petri net-based model that evaluates...

Abstract The paper presents a new low‐complexity algorithm for silence suppression in adverse acoustic environments. uses single time‐domain input parameter (signal power) given to simple matching block. decision module adapts series of thresholds depending on the current estimated signal‐to‐noise‐ratio (SNR) signal. A tests carried out using large speech database confirm 10% improvement pause detection performance as compared with AMR VAD option 1 recently adopted by ETSI 3rd‐generation...

An adaptive system for Voiced/Unvoiced (V/UV) speech detection in the presence of background noise is presented. The results were compared with a non-adaptive classification and V/UV detectors adopted by three important coding standards: LPC10, ITU-T G. 723.1 ETSI AMR. In all cases classifier outperformed traditional solutions.

Tumor infiltrating (TIL) and peripheral blood lymphocytes (PBL) were isolated from 18 patients with non-small cell lung cancer undergoing radical surgery. Surface marker analysis revealed that TILs PBLs mainly consisted of CD3+ T cells generally displayed a lower CD4/CD8 ratio. Differences found in the expression CD25 (IL-2 receptor) DR (MHC class II) antigens, which increased TILs, percentage CD16+ natural killer (NK) cells, was reduced as compared to PBLs. Accordingly, NK activity than...