Logo UNIPVM

ALESSANDRO CURZI

Pubblicazioni

ALESSANDRO CURZI

 

15 pubblicazioni classificate nel seguente modo:

Nr. doc. Classificazioni
8 4 Contributo in Atti di Convegno (Proceeding)
3 1 Contributo su Rivista
2 2 Contributo in Volume
1 5 Altro
1 8 Tesi di dottorato
Anno Risorsa
2015 Speaker Identification with Short Sequences of Speech Frames
Proceedings of the 4th International Conference on Pattern Recognition Applications and Methods (ICPRAM 2015)
Autore/i: Biagetti, Giorgio; Crippa, Paolo; Curzi, Alessandro; Orcioni, Simone; Turchetti, Claudio
Editore: SCITEPRESS (Science and Technology Publications,Lda.)
Classificazione: 4 Contributo in Atti di Convegno (Proceeding)
Abstract: In biometric person identification systems, speaker identification plays a crucial role as the voice is the more natural signal to produce and the simplest to acquire. Mel frequency cepstral coefficients (MFCCs) have been widely adopted for decades in speech processing to capture the speech-specific characteristics with a reduced dimensionality. However, although their ability to de-correlate the vocal source and the vocal tract filter make them suitable for speech recognition, they show up some drawbacks in speaker recognition. This paper presents an experimental evaluation showing that reducing the dimension of features by using the discrete Karhunen-Loève transform (DKLT), guarantees better performance with respect to conventional MFCC features. In particular with short sequences of speech frames, that is with utterance duration of less than 1 s, the performance of truncated DKLT representation are always better than MFCC.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/227759 Collegamento a IRIS

2015 Analysis of the EMG Signal During Cyclic Movements Using Multicomponent AM-FM Decomposition
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
Autore/i: Biagetti, Giorgio; Crippa, Paolo; Curzi, Alessandro; Orcioni, Simone; Turchetti, Claudio
Classificazione: 1 Contributo su Rivista
Abstract: Sport, fitness, as well as rehabilitation activities, often require the accomplishment of repetitive movements. The correctness of the exercises is often related to the capability of maintaining the required cadence and muscular force. Failure to maintain the required force, also known as muscle fatigue, is accompanied by a shift in the spectral content of the surface electromyography (EMG) signal towards lower frequencies. This paper presents a novel approach for simultaneously obtaining exercise repetition frequency and evaluating muscular fatigue, as functions of time, by only using the EMG signal. The mean frequency of the amplitude spectrum (MFA) of the EMG signal, considered as a function of time, is directly related to the dynamics of the movement performed and to the fatigue of the involved muscles. If the movement is cyclic, MFA will display the same pattern and its average will tend to decrease. These two effects have been simultaneously modeled by a two-component AM-FM model based on the Hilbert transform. The method was tested on signals recorded using a wireless system applied to healthy subjects performing dumbbell biceps curls, dumbbell lateral rises, and bodyweight squats. Experimental results show the excellent performance of the proposed technique.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/225251 Collegamento a IRIS

2015 Multi-class ECG beat classification based on a Gaussian mixture model of Karhunen-Loève transform
INTERNATIONAL JOURNAL OF SIMULATION: SYSTEMS, SCIENCE & TECHNOLOGY
Autore/i: Crippa, Paolo; Curzi, Alessandro; Falaschetti, Laura; Turchetti, Claudio
Classificazione: 1 Contributo su Rivista
Abstract: Cardiovascular diseases are one of the main causes of death around the world. Automatic classification of electrocardiogram (ECG) signals is of paramount importance in the unmanned detection of a wide range of heartbeat abnormalities. In this paper an effective multi-class beat classifier, based on a statistical identification of a minimum-complexity model, is presented. This methodology extracts from the ECG signal the multivariate relationships of its natural modes, by means of the separation property of the Karhunen-Loève transform (KLT). Then, it exploits an optimized expectation maximization (EM) algorithm to find the optimal parameters of a Gaussian mixture model, with the focus being in reducing the number of parameters. The resulting statistical model is thus based on the estimation of the multivariate probability density function (PDF) that characterizes each beat type. Based on the above statistical characterization a multi-class ECG classification was performed. The experiments, conducted on the ECG signals from the MIT-BIH arrhythmia database, demonstrated the validity and, considering the reduced model size, the excellent performance of this technique to classify the ECG signals into different disease categories.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/232454 Collegamento a IRIS

2015 Distributed Speech Recognition for Lighting System Control
Intelligent Decision Technologies - Proceedings of the 7th KES International Conference on Intelligent Decision Technologies (KES-IDT 2015)
Autore/i: Biagetti, Giorgio; Crippa, Paolo; Curzi, Alessandro; Falaschetti, Laura; Orcioni, Simone; Turchetti, Claudio
Editore: Springer International Publishing
Luogo di pubblicazione: Heidelberg
Classificazione: 2 Contributo in Volume
Abstract: This paper presents a distributed speech recognition (DSR) system for home/office lighting control by means of users' voice. In this scheme a back-end processes audio signals and transforms them into commands, so that they can be sent to the desired actuators of the lighting system. This paper discusses in detail the solutions and strategies we adopted to improve recognition accuracy and spotting command efficiency in home/office environments, i.e. in situations that involve distant speech and great amounts of background noise or unrelated sounds. Suitable solutions implemented in this recognition engine are able to detect commands also in a continuous listening context and the used DSR strategies greatly simplify the system installation and maintenance. A case study that implements the voice control of a digital addressable lighting interface (DALI) based lighting system has been selected to show the validity and the performance of the proposed system.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/227753 Collegamento a IRIS

2014 A Multi-Class ECG Beat Classifier Based on the Truncated KLT Representation
Proceedings of the 2014 UKSim-AMSS 8th European Modelling Symposium (EMS 2014)
Autore/i: Biagetti, Giorgio; Crippa, Paolo; Curzi, Alessandro; Orcioni, Simone; Turchetti, Claudio
Editore: IEEE Computer Society
Classificazione: 4 Contributo in Atti di Convegno (Proceeding)
Abstract: Automatic classification of electrocardiogram (ECG) signals is of paramount importance in the detection of a wide range of heartbeat abnormalities as aid to improve the diagnostic achieved by cardiologists. In this paper an effective multi-class beat classifier, based on statistical identification of a minimum-complexity model, is proposed. The classifier is trained by extracting from the ECG signal a multivariate random vector by means of a truncated Karhunen-Loève transform (KLT) representation. The resulting statistical model is thus estimated using a robust and efficient Expectation Maximization (EM) algorithm to find the optimal parameters of a Gaussian mixture model. Based on the above statistical characterization a multi-class ECG classifier was derived. The experiments, conducted on the ECG signals from the MIT-BIH arrhythmia database, demonstrated the excellent performance of this technique to classify the ECG signals into different disease categories, with a reduced model complexity.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/224519 Collegamento a IRIS

2014 A Speech Interaction System for an Ambient Assisted Living Scenario
Ambient Assisted Living
Autore/i: M. Alessandrini; G. Biagetti; A. Curzi; C. Turchetti
Editore: Springer International Publishing
Classificazione: 2 Contributo in Volume
Abstract: In this work we describe a speech recognition system aimed at controlling various apparatus of an intelligent home. The system is especially tailored, and ad-hoc optimizations and strategies have been implemented, to make it suitable to operate unobtrusively in the ambient, requiring that the user only installs small and cheap audio front-ends that will capture his spoken commands. A recognition back-end, running either as a network service reached over the Internet or on a PC in the user’s home, performs the hard work of processing the data and turning it into commands, which are sent back to the desired actuator in the home. A case study involving the voice control of a DALI lighting system is presented, together with ideas and results on how to improve recognition accuracy and command spotting efficiency of a system which, by its very nature, might have to deal with audio captured from a distance and great amounts of background noise and unrelated sounds.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/205829 Collegamento a IRIS

2013 Iterative Constrained MLLR Approach for Speaker Adaptation
Proceedings of the 10th IASTED International Conference on Signal Processing, Pattern Recognition, and Applications (SPPRA 2013)
Autore/i: G. Biagetti; A. Curzi; M. Mercuri; C. Turchetti
Classificazione: 4 Contributo in Atti di Convegno (Proceeding)
Abstract: In this paper an effective technique for speaker adaptation on the feature domain is presented. This technique starts from the well known maximum-likelihood linear regression (MLLR) auxiliary function to obtain the constrained MLLR transformation in an iterative fashion. The proposed approach is particularly suitable to be implemented on the client side of a distributed speech recognition scheme, due to the reduced number of iterations required to reach convergence. Extensive experimentation using the CMU Sphinx 4 ASR system along with a preliminarily trained speaker-independent acoustic model for the Italian language, in a setting designed for large-vocabulary continuous speech recognition, demonstrates the effectiveness of the approach even with small amounts of adaptation data.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/97462 Collegamento a IRIS

2013 A Speech Interaction System for an Ambient Assisted Living Scenario
Atti del 4º Forum Italiano per l'Ambient Assisted Living (FORITAAL 2013)
Autore/i: Michele Alessandrini; Giorgio Biagetti; Alessandro Curzi; Claudio Turchetti
Classificazione: 4 Contributo in Atti di Convegno (Proceeding)
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/163548 Collegamento a IRIS

2013 A garbage model generation technique for embedded speech recognisers
Proceedings of the 2013 Signal Processing: Algorithms, Architectures, Arrangements, and Applications (SPA 2013)
Autore/i: Michele Alessandrini; Giorgio Biagetti; Alessandro Curzi; Claudio Turchetti
Classificazione: 4 Contributo in Atti di Convegno (Proceeding)
Abstract: In this paper we present a simple but effective technique to help the designer of a voice-operated appliance add out-of-grammar command rejection capabilities, with a minimal effort and without overly degrading the recognition accuracy. Given the desired operational grammar of the appliance, and starting from a generic pre-trained acoustic model and comprehensive dictionary, we use a speech recogniser to identify suitable decoys to be added to the target grammar. These decoys will capture most of the spoken out-of-vocabulary words, and with appropriate changes to the desired grammar, will make the rejection of unintended commands quite easy. An evaluation of the performance of the proposed approach has been carried out on a sample appliance we developed, and tested with several users, under different acoustic conditions, in a command-spotting scenario. The reported results show that the proposed approach largely outperforms the standard phone loop-based approach.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/158106 Collegamento a IRIS

2011 Semi-automatic acoustic model generation from large unsynchronized audio and text chunks
Proceedings of the 12th Annual Conference of the International Speech Communication Association
Autore/i: Alessandrini M.; Biagetti G.; Curzi A.; Turchetti C.
Classificazione: 4 Contributo in Atti di Convegno (Proceeding)
Abstract: In this paper an effective technique to train an acoustic model from large and unsynchronized audio and text chunks is presented. Given such a speech corpus, an algorithm to automatically segment each chunk into smaller fragments and to synchronize those to the corresponding text is defined. These smaller fragments are more suitable to be used in standard model training algorithms for usage in automatic speech recognition systems. The proposed approach is particularly suitable to bootstrap language models without relying neither on specialized training material nor borrowing from models trained for other similar languages. Extensive experimentation using the CMU Sphinx 4 recognizer and the SphinxTrain model generator in a setting designed for large-vocabulary continuous speech recognition shows the effectiveness of the approach.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/62801 Collegamento a IRIS

2010 Unsupervised identification of nonstationary dynamical systems using a Gaussian mixture model based on EM clustering of SOMs
Proceedings of 2010 IEEE International Symposium on Circuits and Systems (ISCAS)
Autore/i: G. BIAGETTI; P. CRIPPA; A. CURZI; C. TURCHETTI
Editore: IEEE
Luogo di pubblicazione: Piscataway
Classificazione: 4 Contributo in Atti di Convegno (Proceeding)
Abstract: In this paper an effective unsupervised statistical identification technique for nonstationary nonlinear systems is presented. This technique extracts from the system outputs the multivariate relationships of the system natural modes, by means of the separation property of the Karhunen-Loève transform (KLT). Then, it applies a Self-Organizing Map (SOM) to the KLT output vectors in order to give an optimal representation of data. Finally, it exploits an optimized Expectation Maximization (EM) algorithm to find the optimal parameters of a Gaussian mixture model. The resulting statistical system identification is thus based on the estimation of the multivariate probability density function (PDF) of system outputs, whose convergence towards that computed by kernel estimation has also been proved by verifying the asymptotically vanishing of Kullback-Leibler divergences. A large number of simulations on ECG signals demonstrated the validity and the excellent performance of this technique along with its applicability to noninvasive diagnosis of a large class of medical pathologies originated by unknown, unpractical to measure, physiological factors.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/50613 Collegamento a IRIS

2010 Piecewise linear second moment statistical simulation of ICs affected by non-linear statistical effects
INTERNATIONAL JOURNAL OF CIRCUIT THEORY AND APPLICATIONS
Autore/i: BIAGETTI G; P. CRIPPA; CURZI A; ORCIONI S; TURCHETTI C
Classificazione: 1 Contributo su Rivista
Abstract: This paper presents a methodology for statistical simulation of non-linear integrated circuits affected by device mismatch. This simulation technique is aimed at helping designers maximize yield, since it can be orders of magnitude faster than other readily available methods, e.g. Monte Carlo. Statistical analysis is performed by modeling the electrical effects of tolerances by means of stochastic current or voltage sources, which depend on both device geometry and position across the die. They alter the behavior of both linear and non-linear components according to stochastic device models, which reflect the statistical properties of circuit devices up to the second order (i.e. covariance functions). DC, AC, and transient analyses are performed by means of the stochastic modified nodal analysis, using a piecewise linear stochastic technique with respect to the stochastic sources, around a few automatically selected points. Several experimental results on significant circuits, encompassing both the analog and the digital domains, prove the effectiveness of the proposed method.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/51084 Collegamento a IRIS

2009 Tecniche non Monte Carlo per la simulazione statistica di circuiti integrati mixed-signal
Editore: Università Politecnica delle Marche
Classificazione: 8 Tesi di dottorato
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/242276 Collegamento a IRIS

2008 A novel approach to statistical simulation of ICs affected by non-linear variabilities
Proceedings of 2008 IEEE International Symposium on Circuits and Systems
Autore/i: Giorgio, Biagetti; Paolo, Crippa; Alessandro, Curzi; Simone, Orcioni; Claudio, Turchetti
Editore: IEEE
Classificazione: 4 Contributo in Atti di Convegno (Proceeding)
Abstract: This paper presents a methodology for statistical simulation of non-linear integrated circuits affected by device mismatch. This simulation technique is aimed at helping designers maximize yield, since it can be orders of magnitude faster than other readily available methods, e.g. Monte Carlo. DC, AC, and transient analyses are performed by means of the stochastic modified nodal analysis, using a piecewise linearization technique with respect to the stochastic sources, around a few automatically selected points.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/52330 Collegamento a IRIS

2007 SiSMA: Simulator for Statistical Mismatch Analysis
presented at the University Booth at the 10th Design, Automation and Test in Europe (DATE 07)
Autore/i: BIAGETTI G.; ORCIONI S.; CURZI A.; CRIPPA P.; TURCHETTI C.
Classificazione: 5 Altro
Abstract: Presentato a "The University Booth" della conferenza internazionale "10th Design, Automation and Test in Europe (DATE 07)", 16-20 Aprile, 2007, Nizza, Francia.
Scheda della pubblicazione: https://iris.univpm.it/handle/11566/50212 Collegamento a IRIS


Università Politecnica delle Marche

P.zza Roma 22, 60121 Ancona
Tel (+39) 071.220.1, Fax (+39) 071.220.2324
P.I. 00382520427