Search

Nuisance Alarm Rate (NAR) is critical in $\phi $ -OTDR perturbation detection systems. We present in this letter a novel matched filtering-based feature extractor which aims to noise reduction so that the detection system gets improved performance. This feature extractor requires a small number of data vectors to be acquired which is combined with a random forest-based machine learning strategy to significantly reduce the NAR. In addition, since the number of data vectors is small, this system can also be useful for time-sensitive detection applications.

This paper presents a novel pipeline integrity surveillance system aimed to the detection and classification of threats in the vicinity of a long gas pipeline. The sensing system is based on phase-sensitive optical time domain reflectometry ( $\phi$ -OTDR) technology for signal acquisition and pattern recognition strategies for threat identification. The proposal incorporates contextual information at the feature level in a Gaussian Mixture Model-Hidden Markov Model (GMM-HMM)-based pattern classification system and applies a system combination strategy for acoustic trace decision. System combination relies on majority voting of the decisions given by the individual contextual information sources and the number of states used for HMM modelling. The system runs in two different modes: (1) machine+activity identification, which recognizes the activity being carried out by a certain machine, and (2) threat detection, aimed to detect threats no matter what the real activity being conducted is. In comparison with the previous systems based on the same rigorous experimental setup, the results show that the system combination from the contextual feature information and the GMM-HMM approach improves the results for both machine+activity identification (7.6% of relative improvement with respect to the best published result in the literature on this task) and threat detection (26.6% of relative improvement in the false alarm rate with 2.1% relative reduction in the threat detection rate)., European Commission, Ministerio de Economía y Competitividad, Comunidad de Madrid

In order to promote education in environmental sustainability at higher education institutions and establish action priorities, a detailed analysis of the initial situation is needed, both in terms of courses offered and teaching and assessment strategies implemented. This article presents a methodology that can be used to standardize the evaluation of teaching in environmental sustainability at different universities. We exemplify its application at the University of Alcalá, located in Central Spain. The inventory was conducted using in-house software development for the extraction of environmental sustainability concepts in the degree courses’ syllabi, completed with a survey to faculty members and a manual review to confirm its adjustment to environmental issues. Those finally selected were analyzed in depth. The main results indicate that currently, only a small part of the courses (5.5%) offered at the University explicitly include environmental sustainability content. The grades that concentrate most of the courses are those of Environmental Sciences, Biology, Economics, Tourism and Pharmacy, with very low occurrence in the grades of Education and Health Sciences. It is concluded that further institutional commitment is needed in the definition of medium-term strategies to guide the training and dissemination efforts in order to promote sound environmental education in university courses.

19 pags., 4 figs., 9 tabs. -- This article belongs to the Special Issue Pattern Recognition and Applications, We present a new pipeline integrity surveillance system for long gas pipeline threat detection and classification. The system is based on distributed acoustic sensing with phase-sensitive optical time domain reflectometry (ϕ-OTDR) and pattern recognition for event classification. The proposal incorporates a multi-position approach in a Gaussian Mixture Model (GMM)-based pattern classification system which operates in a real-field scenario with a thorough experimental procedure. The objective is exploiting the availability of vibration-related data at positions nearby the one actually producing the main disturbance to improve the robustness of the trained models. The system integrates two classification tasks: (1) machine + activity identification, which identifies the machine that is working over the pipeline along with the activity being carried out, and (2) threat detection, which aims to detect suspicious threats for the pipeline integrity (independently of the activity being carried out). For the machine + activity identification mode, the multi-position approach for model training obtains better performance than the previously presented single-position approach for activities that show consistent behavior and high energy (between 6% and 11% absolute) with an overall increase of 3% absolute in the classification accuracy. For the threat detection mode, the proposed approach gets an 8% absolute reduction in the false alarm rate with an overall increase of 4.5% absolute in the classification accuracy., This research was funded by a private research project by Fluxys, Gassco and Statoil named “PIT-STOP (Early detection of Pipeline Integrity Threats using a SmarT fiber-OPtic surveillance system)”, by the Spanish Ministry of Economy and Competitiveness under projects HEIMDALUAH (TIN2016-75982-C2-1-R) and ARTEMISA (TIN2016-80939-R), by the University of Alcalá under projects ACUFANO (CCG19/IA-024) and ARGOS (CCG20/IA-043), and by the Spanish Ministry of Science, Innovation and Universities (IJCI-2017-33856 and RTI2018-097957-B-C33). The APC was funded by the Ministry of Science, Innovation and Universities of Spain grant number RTI2018-095324-B-I00.

Smart spaces are environments equipped with a set of sensors with the main objective of understanding humans’ behavior within them, their interactions and to improve human-machine interfaces. Audiovisual tracking is used to know people’s position in the environment and if they are talking, through the use of cameras and microphones. In this work we present an audiovisual tracking solution with a single camera and microphone array in a distributed configuration. Our idea is to exploit the estimation of azimuth and elevation from audio information to be fused with the position estimation obtained from a Viola and Jones based observation model. The fact that the microphone array is not co-located with the camera will allow to reduce the distance estimation uncertainty from the video model and improve tracking accuracy. The system was evaluated on the AV16.3 database on single speaker sequences, outperforming results of state-of-the-art, under these conditions.

We report a novel open competition aimed at evaluating accurate robot position tracking in indoor environments. The competition was organized within the IPIN 2016 (Indoor Positioning and Indoor Navigation international Conference). Here, we describe the competition, the competitors and their final results. The challenges of this new competition included: tracking an industrial robot following an unknown path but with a defined ground-truth, and open positioning system to be deployed on-site, with no restrictions apart from those related to safety issues. Our aim here is to provide sufficient detail to serve as a solid basis for future competition initiatives with a similar scope, using common metrics and objective evaluation procedures. In addition, the real systems evaluated represent state-of-the-art performance, and thus offer interesting solutions to the problem posed in the competition.

This paper proposes a deep learning-based method to detect multiple people from a single overhead depth image with high precision. Our neural network, called DPDnet , is composed by two fully-convolutional encoder-decoder blocks built with residual layers. The main block takes a depth image as input and generates a pixel-wise confidence map, where each detected person in the image is represented by a Gaussian-like distribution, The refinement block combines the depth image and the output from the main block, to refine the confidence map. Both blocks are simultaneously trained end-to-end using depth images and ground truth head position labels. The paper provides a rigorous experimental comparison with some of the best methods of the state-of-the-art, being exhaustively evaluated in different publicly available datasets. DPDnet proves to outperform all the evaluated methods with statistically significant differences, and with accuracies that exceed 99%. The system was trained on one of the datasets (generated by the authors and available to the scientific community) and evaluated in the others without retraining, proving also to achieve high accuracy with varying datasets and experimental conditions. Additionally, we made a comparison of our proposal with other CNN-based alternatives that have been very recently proposed in the literature, obtaining again very high performance. Finally, the computational complexity of our proposal is shown to be independent of the number of users in the scene and runs in real time using conventional GPUs.

This paper describes a novel DNN-based system, named PD3net , that detects multiple people from a single depth image, in real time. The proposed neural network processes a depth image and outputs a likelihood map in image coordinates, where each detection corresponds to a Gaussian-shaped local distribution, centered at each person’s head. This likelihood map encodes both the number of detected people as well as their position in the image, from which the 3D position can be computed. The proposed DNN includes spatially separated convolutions to increase performance, and runs in real-time with low budget GPUs. We use synthetic data for initially training the network, followed by fine tuning with a small amount of real data. This allows adapting the network to different scenarios without needing large and manually labeled image datasets. Due to that, the people detection system presented in this paper has numerous potential applications in different fields, such as capacity control, automatic video-surveillance, people or groups behavior analysis, healthcare or monitoring and assistance of elderly people in ambient assisted living environments. In addition, the use of depth information does not allow recognizing the identity of people in the scene, thus enabling their detection while preserving their privacy. The proposed DNN has been experimentally evaluated and compared with other state-of-the-art approaches, including both classical and DNN-based solutions, under a wide range of experimental conditions. The achieved results allows concluding that the proposed architecture and the training strategy are effective, and the network generalize to work with scenes different from those used during training. We also demonstrate that our proposal outperforms existing methods and can accurately detect people in scenes with significant occlusions.

One of the most popular techniques for Acoustic Source Localization is the Generalized Cross Correlation (GCC) and its use in Steered Response Power techniques (SRP). Nowadays, Deep Learning strategies may outperform these classical methods, but they are generally dependent on the room and sensor geometric configuration that are used during the training phases. Hence, they require adaptation and re-training when facing a new environment, which is a problem in practice as re-training requires labelling new data and running a complex training algorithm. In this work we use a Convolutional Deep Neural Network that transforms the GCC between two signals into a Gaussian shaped signal, that we call Deep Generalized Cross Correlation (DeepGCC). We combine DeepGCC estimations to create a 3D acoustic map, similarly to SRP techniques. This acoustic map can be further refined using a sparse generative model to recover the source position. Crucially, we can adapt the acoustic map to different microphone array geometries without retraining the DeepGCC network. We show that our method outperforms both classical approaches and recent Deep Learning strategies in real and simulated challenging scenarios with mismatched training-testing conditions, not requiring re-training with different sensor configurations or room environments.

Robust system to detect people only using depth information from a ToF camera.Refined algorithm for determining the regions of interest.Classifier stage to distinguish between people and other objects in the scene.Rigorous experimental procedure shows high performance with a broad user variability.Generated database will be made available to the research community. In this paper we describe a system for the automatic detection of multiple people in a scene, by only using depth information provided by a Time of Flight (ToF) camera placed in overhead position. The main contribution of this work lies in the proposal of a methodology for determining the Regions of Interest (ROI's) and feature extraction, which result in a robust discrimination between people with or without accessories and objects (either static or dynamic), even when people and objects are close together. Since only depth information is used, the developed system guarantees users' privacy. The designed algorithm includes two stages: an online stage, and an offline one. In the offline stage, a new depth image dataset has been recorded and labeled, and the labeled images have been used to train a classifier. The online stage is based on robustly detecting local maximums in the depth image (which are candidates to correspond to the head of the people present in the scene), from which a carefully ROI is defined around each of them. For each ROI, a feature vector is extracted, providing information on the top view of people and objects, including information related to the expected overhead morphology of the head and shoulders. The online stage also includes a pre-filtering process, in order to reduce noise in the depth images. Finally, there is a classification process based on Principal Components Analysis (PCA). The online stage works in real time at an average of 150 fps. In order to evaluate the proposal, a wide experimental validation has been carried out, including different number of people simultaneously present in the scene, as well as people with different heights, complexions, and accessories. The obtained results are very satisfactory, with a 3.1% average error rate.

ϕ-OTDR perturbation detection applications demand optimal precision of the perturbation location. Strategies for improving both signal-to-noise (SNR) and precision of the perturbation location in a laboratory environment may fail when applying to a very long fiber under test (FUT) in real-field environments. With this deployment, meaningful energy points representing the response of a certain perturbation can be located at random locations of the fiber other than the original location of the perturbation. These random locations are referred to as the ghost energy points that confuse the system to mistakenly consider the location of these points as the original perturbation location. We present in this paper a novel space-time scanning (ST-scan) method that segregates the ghost energy point locations from those of the real perturbation so that the original perturbation location estimation is improved.

Time delay estimation is essential in Acoustic Source Localization (ASL) systems. One of the most used techniques for this purpose is the Generalized Cross Correlation (GCC) between a pair of signals and its use in Steered Response Power (SRP) techniques, which estimate the acoustic power at a specific location. Nowadays, Deep Learning strategies may outperform these methods. However, they are generally dependent on the geometric and sensor configuration conditions that are available during the training phases, thus having limited generalization capabilities when facing new environments if no re-training nor adaptation is applied. In this work, we propose a method based on an encoder-decoder CNN architecture capable of outperforming the well known SRP-PHAT algorithm, and also other Deep Learning strategies when working in mismatched training-testing conditions without requiring a model re-training. Our proposal aims to estimate a smoothed version of the correlation signals, that is then used to generate a refined acoustic power map, which leads to better performance on the ASL task. Our experimental evaluation uses three publicly available realistic datasets and provides a comparison with the SRP-PHAT algorithm and other recent proposals based on Deep Learning.

This work proposes a new method for face and mouth tracking using video cameras. This method has been proposed to be used in intelligent spaces as an initial stage to provide information to higher level applications. The method incorporates the classifiers of the Viola and Jones detection method, modified to provide a probabilistic output previously proposed by the authors, useful for tracking with particle filter. The method combines classifiers trained to detect specific poses (frontal and side face views), building an independent likelihood model of pose changes. We also propose to combine the model in several cameras to allow the mouth tracking in a three-dimensional space. The system has been evaluated on the AV16.3 database sequences showing good results in both precision and recall when using a single camera (in a bidimensional space), and an error below 3cm when using three cameras (in a three-dimensional space).

Measuring the Time delay of Arrival (TDOA) between a set of sensors is the basic setup for many applications, such as localization or signal beamforming. This paper presents the set of TDOA matrices, which are built from noise-free TDOA measurements, not requiring knowledge of the sensor array geometry. We prove that TDOA matrices are rank-two and have a special SVD decomposition that leads to a compact linear parametric representation. Properties of TDOA matrices are applied in this paper to perform denoising, by finding the TDOA matrix closest to the matrix composed with noisy measurements. The paper shows that this problem admits a closed-form solution for TDOA measurements contaminated with Gaussian noise which extends to the case of having missing data. The paper also proposes a novel robust denoising method resistant to outliers, missing data and inspired in recent advances in robust low-rank estimation. Experiments in synthetic and real datasets show significant improvements of the proposed denoising algorithms in TDOA-based localization, both in terms of TDOA accuracy estimation and localization error.

This paper presents the first available report in the literature of a system aimed at the detection and classification of threats in the vicinity of a long gas pipeline. The system is based on phase-sensitive optical time-domain reflectometry technology for signal acquisition and pattern recognition strategies for threat identification. The system operates in two different modes: 1) machine+activity identification, which outputs the activity being carried out by a certain machine; and 2) threat detection, aimed at detecting threats no matter what the real activity being conducted is. Different strategies dealing with position selection and normalization methods are presented and evaluated using a rigorous experimental procedure on realistic field data. Experiments are conducted with eight machine+activity pairs, which are further labeled as threat or nonthreat for the second mode of the system. The results obtained are promising given the complexity of the task and open the path to future improvements toward fully functional pipeline threat detection systems operating in real conditions., European Commission, Comunidad de Madrid, Ministerio de Economía y Competitividad, Universidad de Alcalá

Time and spatial domains in ϕ -OTDR perturbation detection and recognition for pipeline and border security applications in very long fiber under test (FUT) environments have not been properly analyzed so far. We propose in this paper several issues and the corresponding solutions that should be considered in both domains when developing those applications. Solutions are based on: (1) the importance of differential signals in direct detected ϕ -OTDR systems; (2) dealing with the useless data acquired in the time domain; (3) dealing with the zero and ghost energy points along the FUT which will undoubtedly lead to hiding the real perturbation and reducing the potential of finding the exact location of the perturbation, respectively. Experimental results show that: (1) setting the appropriate experimental conditions (i.e., increasing the intensity of the perturbation and sampling at the minimum sampling rate that satisfies the Nyquist criterion) is a must; (2) the Kendall correlation-based technique on differential signals allows obtaining more significant data traces both in time and spatial domains; (3) combining both methods obtains further gains; (4) S-Scan algorithm accurately provides the time instant of the real perturbation by increasing the probability of detecting zero energy points and avoiding ghost energy points in further processing.

The algorithms for acoustic source localization based on PHAT filtering have been profusely used with good results in reverberant and noisy environments. However, there are very few studies that give a formal explanation of their robustness, most of them providing just an empirical validation or showing results on simulated data. In this work we present a novel analytical model for predicting the behavior of both the SRP-PHAT power maps and the GCC-PHAT functions. The results show that they are only affected by the signal bandwidth, the microphone array topology, and the room geometry, being independent of the spectral content of the received signal. The proposed model is shown to be valid in reverberant environments and under far and near field conditions. Using this result, an analysis study on how the aforementioned factors affect the SRP-PHAT power maps is presented providing well supported theoretical and practical considerations. The model validation is based on both synthetic and real data, obtaining in all cases a high accuracy of the model to reproduce the SRP-PHAT power maps, both in anechoic and non-anechoic scenarios, becoming thus an excellent tool to be exploited for the improvement of real world relevant applications related to acoustic localization. HighlightsA novel parametric analytical model to predict SRP-PHAT power maps is formulated.An exhaustive evaluation is done on both synthetic and real data.Results show high accuracy for very different acoustical and geometrical conditions.The paper also addresses practical issues in the model implementation.

This paper presents an on-line augmented surveillance system that aims to real time monitoring of activities along a pipeline. The system is deployed in a fully realistic scenario and exposed to real activities carried out in unknown places at unknown times within a given test time interval (socalled blind field tests). We describe the system architecture that includes specific modules to deal with the fact that continuous on-line monitoring needs to be carried out, while addressing the need of limiting the false alarms at reasonable rates. To the best or our knowledge, this is the first published work in which a pipeline integrity threat detection system is deployed in a realistic scenario (using a fiber optic along an active gas pipeline) and is thoroughly and objectively evaluated in realistic blind conditions. The system integrates two operation modes: The machine+activity identification mode identifies the machine that is carrying out a certain activity along the pipeline, and the threat detection mode directly identifies if the activity along the pipeline is a threat or not. The blind field tests are carried out in two different pipeline sections: The first section corresponds to the case where the sensor is close to the sensed area, while the second one places the sensed area about 35 km far from the sensor. Results of the machine+activity identification mode showed an average machine+activity classification rate of 46:6%. For the threat detection mode, 8 out of 10 threats were correctly detected, with only 1 false alarm appearing in a 55:5-hour sensed period., European Commission, Ministerio de Economía y Competitividad, Comunidad de Madrid

25th International Conference on Optical Fiber Sensors, 2017, Jeju, Corea del Sur., This paper presents the first report on on-line and final blind field test results of a pipeline integrity threat surveillance system. The system integrates a machine+activity identification mode, and a threat detection mode. Two different pipeline sections were selected for the blind tests: One close to the sensor position, and the other 35 km away from it. Results of the machine+activity identification mode showed that about 46% of the times the machine, the activity or both were correctly identified. For the threat detection mode, 8 out of 10 threats were correctly detected, with 1 false alarm., European Commission, Ministerio de Economía y Competitividad, Comunidad de Madrid

This paper presents a novel surveillance system aimed at the detection and classification of threats in the vicinity of a long gas pipeline. The sensing system is based on phase-sensitive optical time domain reflectometry (φ-OTDR) technology for signal acquisition and pattern recognition strategies for threat identification. The proposal incorporates contextual information at the feature level and applies a system combination strategy for pattern classification. The contextual information at the feature level is based on the tandem approach (using feature representations produced by discriminatively-trained multi-layer perceptrons) by employing feature vectors that spread different temporal contexts. The system combination strategy is based on a posterior combination of likelihoods computed from different pattern classification processes. The system operates in two different modes: (1) machine + activity identification, which recognizes the activity being carried out by a certain machine, and (2) threat detection, aimed at detecting threats no matter what the real activity being conducted is. In comparison with a previous system based on the same rigorous experimental setup, the results show that the system combination from the contextual feature information improves the results for each individual class in both operational modes, as well as the overall classification accuracy, with statistically-significant improvements., European Commission, Ministerio de Economía y Competitividad, Comunidad de Madrid

Optimal placement of sensors is a fundamental task to achieve high performance of different applications, especially in multi-sensor scenarios and complex environments. This paper proposes an alternative for automatically estimating the optimal placement of multiple microphones (individual or conforming arrays), in arbitrary complex indoor environments, for acoustic localization tasks. The proposal is based on evolutionary optimization strategies (genetic algorithms). We also propose a spatial likelihood function based on a recently described acoustic model which is able to accurately predict an acoustic power map for a given sensor and source configuration. The spatial likelihood function is the base to develop error metrics, from which different objective performance measures are derived. An extensive evaluation has been carried out, generating different placement alternatives according to the number of microphones, their arrangement (individual or conforming arrays), and the optimization criteria used in the optimization process. In addition, these results have been compared with the ones obtained by manual and random sensor placement, showing that the proposal clearly outperforms them, thus allowing to validate it.

This paper presents a novel approach for indoor acoustic source localization using sensor arrays. The proposed solution starts by defining a generative model, designed to explain the acoustic power maps obtained by Steered Response Power (SRP) strategies. An optimization approach is then proposed to fit the model to real input SRP data and estimate the position of the acoustic source. Adequately fitting the model to real SRP data, where noise and other unmodelled effects distort the ideal signal, is the core contribution of the paper. Two basic strategies in the optimization are proposed. First, sparse constraints in the parameters of the model are included, enforcing the number of simultaneous active sources to be limited. Second, subspace analysis is used to filter out portions of the input signal that cannot be explained by the model. Experimental results on a realistic speech database show statistically significant localization error reductions of up to 30% when compared with the SRP-PHAT strategies.

Several issues concerning the current use of speech interfaces are discussed and the design and development of a speech interface that enables air traffic controllers to command and control their terminals by voice is presented. A special emphasis is made in the comparison between laboratory experiments and field experiments in which a set of ergonomics-related effects are detected that cannot be observed in the controlled laboratory experiments. The paper presents both objective and subjective performance obtained in field evaluation of the system with student controllers at an air traffic control (ATC) training facility. The system exhibits high word recognition test rates (0.4% error in Spanish and 1.5% in English) and low command error (6% error in Spanish and 10.6% error in English in the field tests). Subjective impression has also been positive, encouraging future development and integration phases in the Spanish ATC terminals designed by Aeropuertos Espanoles y Navegacion Aerea (AENA).

This paper describes a proposal for a full body gesture recognition system to be used in an intelligent space to allow users to control their environment. We describe a successful adaptation of the traditional strategy applied in the design of spoken language recognition systems, to the new domain of full body gesture recognition. The experimental evaluation has been done on a realistic task where different elements in the environment can be controlled by the users using gesture sequences. The evaluation results have been obtained applying a rigorous experimental procedure, evaluating different feature extraction strategies. The average recognition rates achieved are around 97 % for the gestural sentence level, and over 98 % at the gesture level, thus experimentally validating the proposal.

Fundamento y objetivo El proposito de este trabajo es investigar los factores predictores de mortalidad y la estancia media de los pacientes diagnosticados de angina inestable (AI) durante su ingreso en las unidades de cuidados intensivos o unidades de cuidados coronarios (UCI/UCC). Pacientes y metodo Estudio de cohorte retrospectivo que incluyo a los pacientes del registro espanol ARIAM (Analisis del Retraso en el Infarto Agudo de Miocardio). El periodo de estudio comprendio de junio de 1996 a diciembre de 2003. Su seguimiento se limito a UCI/UCC. Se realizo un analisis univariante entre los pacientes fallecidos y los vivos y otro entre los pacientes con estancia prolongada y no prolongada. Se realizaron ademas 3 analisis multivariantes con el objetivo de evaluar los factores relacionados con la mortalidad, las variables asociadas a la realizacion de intervencion coronaria percutanea (ICP) y los factores asociados a la estancia media prolongada en UCI/UCC. Resultados Se incluyo a 14.096 pacientes con AI. La mortalidad fue del 1,1% y se asocio a la clase Killip, edad, necesidad de reanimacion cardiopulmonar y presencia de shock cardiogenico,arritmias y diabetes; mientras que los pacientes fumadores se asociaron a una menor mortalidad. Se realizo ICP solo a 1.226 (8,9%) pacientes, aunque su realizacion se incremento con los anos. La ICP se asocio a edad, traslado interhospitalario, tabaquismo, infarto de miocardio previo, shock cardiogenico o ante el bloqueo auriculoventricular de alto grado (BAVAG) y tratamiento con bloqueadores beta. La estancia media (desviacion estandar) fue de 3,15 (18,65) dias. La estancia media prolongada dependio de la edad, haber sido sometido a una coronariografia o ICP, clase Killip,ecocardiografia o ventilacion mecanica y presentar complicaciones graves. Conclusiones La mortalidad se explica por la comorbilidad y la gravedad de la isquemia. Los pacientes a los que se realiza ICP son una poblacion menos grave, y su realizacion con el tiempo se incrementa. La estancia media prolongada se asocia a la mayor gravedad de la isquemia miocardica y a la ICP.

24th International Conference on Optical Fibre Sensors, Curitiba, Brazil, September 28, 2015., The preliminary results of a surveillance system set up for real time monitoring activities along a pipeline and analyzing for possible threats are presented. The system consists of a phi-OTDR based sensor used to monitor vibrations along an optical fiber combined with a pattern recognition system that classifies the recorded signals. The acoustic traces generated by the activities of different machines at various locations along a pipeline were recorded in the field. The signals, corresponding to machinery activities, were clearly distinguished from background noise. A threat classification rate of 68.11% with 55.55% false alarms was obtained., European Commission, Ministerio de Economía y Competitividad, Comunidad de Madrid, Universidad de Alcalá, Ministerio de Ciencia e Innovación

We present an overview of the most common techniques used in automatic speech recognition to adapt a general system to a different environment (known as cross-task adaptation) such as in an air traffic control system (ATC). The conditions present in ATC are very specific: very spontaneous, the presence of noise, and high speed speech. So, with a typical speech recognizer the recognition results are unsatisfactory. We have to decide on the best option for the modeling: to develop acoustic models specific to those conditions from scratch using the data available for the new environment, or to carry out cross-task adaptation starting from reliable HMM models (usually requiring less data in the target domain). We begin with a description of the main techniques considered for cross-task adaptation, namely maximum a posteriori (MAP), maximum likelihood linear regression (MLLR), and the two together. We have applied each in two speech recognizers for air traffic control tasks, one for spontaneous speech and the other for a command interface. We show the performance of these techniques and compare them with the development of a new system from scratch. We also show the results obtained for speaker adaptation using a variable amount of adaptation data. The main conclusion is that MLLR can outperform MAP when a large number of transforms is used, and MLLR followed by MAP is the best option. All of these techniques are better than developing a new system from scratch, showing the effectiveness of mean and variance adaptation

This paper presents an approach to design Electronic Systems Curricula for making electronics more appealing to students. Since electronics is an important grounding for other disciplines (computer science, signal processing, and communications), this approach proposes the development of multidisciplinary projects using the project-based learning (PBL) strategy for increasing the attractiveness of the curriculum. The proposed curriculum structure consists of eight courses: four theoretical courses and four PBL courses (including a compulsory Master's thesis). In PBL courses, the students, working together in groups, develop multidisciplinary systems, which become progressively more complex. To address this complexity, the Department of Electronic Engineering has invested in the last five years in many resources for developing software tools and a common hardware. This curriculum has been evaluated successfully for the last four academic years: the students have increased their interest in electronics and have given the courses an average grade of more than 71% for all PBL course evaluations (data extracted from students surveys). The students have also acquired new skills and obtained very good academic results: the average grade was more than 74% for all PBL courses. An important result is that all students have developed more complex and sophisticated electronic systems, while considering that the results are worth the effort invested

This article describes a search algorithm adapted to the Spanish Railway Network for generating as many traveling options as possible between two railway stations. This algorithm (Warshall's algorithm) uses connecting matrices to find all possible railway journeys. The Spanish Railway Company has imposed severe restrictions: less than 1 second per query in a 600Mhz processor PC with 32Mb RAM and 150Mb hard disk free memory. The final average time for a simple query is around 0.25 seconds and the whole memory consumption is 127Mb. The final implementation has been divided into 3 modules. In the first module, we store additional information in the connecting matrices to accelerate the later search, proposing several strategies for reducing thier size. The journey option calculation module accesses the matrix information and composes the traveling options. Finally, in the filtering module we describe the selection criteria considering the algorithm embedded in a general information service.

In this paper, we propose a strategy for designing dialogue managers in spoken dialogue systems for a restricted domain. This strategy combines several information sources intuition, observation and simulation, in order to maximize the adaptation within the system capability and the expectation of the user. These sources are combined by an iterative process consisting of five steps, where different dialogue alternatives are proposed and evaluated sequentially. The evaluation process includes different measures depending on the information required. Several measures are proposed and analyzed in each step. We also describe a user-modeling technique and an approach for designing the confirmation sub-dialogues based on recognition confidence measures. The knowledge-combining methodology is described and applied to a railway information system. In a subjective evaluation, users from the university gave the system a 3.9 score on a 5-point scale with an average call duration of 205 seconds. The employers of the railway company were more critical of the system. They gave it a score of 2.1 even though the system resolved more than half of the calls (57.8%) within an average call duration of three minutes (185 seconds).

The Viola and Jones face detectors and Particle Filters are great algorithms for face detections and target tracking. However Viola outputs a binary result, while Particle Filters work with probabilistic inputs. This is the reason why there are not so many works that combine both algorithms. A probabilistic model or likelihood functions to transform Viola and Jones output to probabilistic data are needed to allow linking both methods. In this work we explore some Viola and Jones based likelihood functions presented in literature, and propose new strategies. We also extend the evaluation of the likelihood functions in position, scale and pose. One of our proposed functions shows better characteristics to be used in intelligent spaces in three dimensional face tracking applications.

Although there has been a lot of interest in recognizing and understanding air traffic control (ATC) speech, none of the published works have obtained detailed field data results. We have developed a system able to identify the language spoken and recognize and understand sentences in both Spanish and English. We also present field results for several in-tower controller positions. To the best of our knowledge, this is the first time that field ATC speech (not simulated) is captured, processed, and analyzed. The use of stochastic grammars allows variations in the standard phraseology that appear in field data. The robust understanding algorithm developed has 95% concept accuracy from ATC text input. It also allows changes in the presentation order of the concepts and the correction of errors created by the speech recognition engine improving it by 17% and 25%, respectively, absolute in the percentage of fully correctly understood sentences for English and Spanish in relation to the percentages of fully correctly recognized sentences. The analysis of errors due to the spontaneity of the speech and its comparison to read speech is also carried out. A 96% word accuracy for read speech is reduced to 86% word accuracy for field ATC data for Spanish for the "clearances" task confirming that field data is needed to estimate the performance of a system. A literature review and a critical discussion on the possibilities of speech recognition and understanding technology applied to ATC speech are also given.

We have applied two state-of-the-art speech synthesis techniques (unit selection and HMM-based synthesis) to the synthesis of emotional speech. A series of carefully designed perceptual tests to evaluate speech quality, emotion identification rates and emotional strength were used for the six emotions which we recorded - happiness, sadness, anger, surprise, fear, disgust. For the HMM-based method, we evaluated spectral and source components separately and identified which components contribute to which emotion.Our analysis shows that, although the HMM method produces significantly better neutral speech, the two methods produce emotional speech of similar quality, except for emotions having context-dependent prosodic patterns. Whilst synthetic speech produced using the unit selection method has better emotional strength scores than the HMM-based method, the HMM-based method has the ability to manipulate the emotional strength. For emotions that are characterized by both spectral and prosodic components, synthetic speech using unit selection methods was more accurately identified by listeners. For emotions mainly characterized by prosodic components, HMM-based synthetic speech was more accurately identified. This finding differs from previous results regarding listener judgements of speaker similarity for neutral speech. We conclude that unit selection methods require improvements to prosodic modeling and that HMM-based methods require improvements to spectral modeling for emotional speech. Certain emotions cannot be reproduced well by either method.

In this paper a Bayesian Networks, BNs, approach to dialogue modelling [1] is evaluated in terms of a battery of both subjective and objective metrics. A significant effort in improving the contextual information handling capabilities of the system has been done. Consequently, besides typical dialogue measurement rates for usability like task or dialogue completion rates, dialogue time, etc. we have included a new figure measuring the contextuality of the dialogue as the number of turns where contextual information is helpful for dialogue resolution. The evaluation is developed through a set of predefined scenarios according to different initiative styles and focusing on the impact of the user’s level of experience.

This article describes an architecture for translating speech into Spanish Sign Language (SSL). The architecture proposed is made up of four modules: speech recognizer, semantic analysis, gesture sequence generation and gesture playing. For the speech recognizer and the semantic analysis modules, we use software developed by IBM and CSLR (Center for Spoken Language Research at University of Colorado), respectively. Gesture sequence generation and gesture animation are the modules on which we have focused our main effort. Gesture sequence generation uses semantic concepts (obtained from the semantic analysis) associating them with several SSL gestures. This association is carried out based on a number of generation rules. For gesture animation, we have developed an animated agent (virtual representation of a human person) and a strategy for reducing the effort in gesture animation. This strategy consists of making the system automatically generate all agent positions necessary for the gesture animation. In this process, the system uses a few main agent positions (two or three per second) and some interpolation strategies, both issues previously generated by the service developer (the person who adapts the architecture proposed in this paper to a specific domain). Related to this module, we propose a distance between agent positions and a measure of gesture complexity. This measure can be used to analyze the gesture perception versus its complexity. With the architecture proposed, we are not trying to build a domain independent translator but a system able to translate speech utterances into gesture sequences in a restricted domain: railway, flights or weather information.

This paper describes the development of and the first experiments in a Spanish to sign language translation system in a real domain. The developed system focuses on the sentences spoken by an official when assisting people applying for, or renewing their Identity Card. The system translates official explanations into Spanish Sign Language (LSE: Lengua de Signos Espan¿ola) for Deaf people. The translation system is made up of a speech recognizer (for decoding the spoken utterance into a word sequence), a natural language translator (for converting a word sequence into a sequence of signs belonging to the sign language), and a 3D avatar animation module (for playing back the hand movements). Two proposals for natural language translation have been evaluated: a rule-based translation module (that computes sign confidence measures from the word confidence measures obtained in the speech recognition module) and a statistical translation module (in this case, parallel corpora were used for training the statistical model). The best configuration reported 31.6% SER (Sign Error Rate) and 0.5780 BLEU (BiLingual Evaluation Understudy). The paper also describes the eSIGN 3D avatar animation module (considering the sign confidence), and the limitations found when implementing a strategy for reducing the delay between the spoken utterance and the sign sequence animation.

In this paper we describe recent work oriented towards establishing some background to support the design of successful "universal" (language independent, specifically) emotional speech identification systems. Our final goal is using such systems to improve state of the art advanced spoken language systems. We carried out several experiments with emotional speech databases for Spanish (SES) and German (EMODB), including language-dependent tasks (to establish the accuracy of the identification systems), and language independent (cross-language) tasks. We show how automatic emotion identification results can be comparable to those obtained by human listeners, provided that enough training data is available. Moreover, and in spite of data availability, we also show how there are emotions (sadness and anger) that are clearly identified in the automatic experiments, for both languages, showing a reasonably clear "language-independent" behavior of such emotions. Finally, we show how psychological considerations related to the emotional speech generation process can be somehow related to our experimental results, suggesting that previously identifying such psychological features may greatly help the development of high quality "universal" emotional speech identification systems.