Matthieu VOIRY - Academia.edu (original) (raw)
Papers by Matthieu VOIRY
In this paper we investigate into perceptual properties of Style-WaveGAN, a drum synthesis method... more In this paper we investigate into perceptual properties of Style-WaveGAN, a drum synthesis method proposed in a previous publication. For both, the sound quality as well as the control precision StyleWaveGAN has been shown to deliver state of the art performance for quantitative metrics (FAD and MSE of the control parameters). The present paper aims to provide insight into the perceptual relevance of these results. Accordingly, we performed a subjective evaluation of the sound quality as well as a subjective evaluation of the precision of the control using timbre descriptors from the AudioCommons toolbox. We evaluate the sound quality with mean opinion score and make measurements of psychophysical response to the variations of the control. By means of the perceptual tests, we demonstrate that StyleWaveGAN produces better sound quality than state-of-the-art model DrumGAN and that the mean control error is lower than the absolute threshold of perception at every point of measurement used in the experiment.
Dans différents domaines industriels, la problématique du diagnostic prend une place importante. ... more Dans différents domaines industriels, la problématique du diagnostic prend une place importante. Ainsi, le contrôle d’aspect des composants optiques est une étape incontournable pour garantir leurs performances opérationnelles. La méthode conventionnelle de contrôle par un opérateur humain souffre de limitations importantes qui deviennent insurmontables pour certaines optiques hautes performances. Dans ce contexte, cette thèse traite de la conception d’un système automatique capable d’assurer le contrôle d’aspect. Premièrement, une étude des capteurs pouvant être mis en oeuvre par ce système est menée. Afin de satisfaire à des contraintes de temps de contrôle, la solution proposée utilise deux capteurs travaillant à des échelles différentes. Un de ces capteurs est basé sur la microscopie Nomarski ; nous présentons ce capteur ainsi qu’un ensemble de méthodes de traitement de l’image qui permettent, à partir des données fournies par celui-ci, de détecter les défauts et de déterminer l...
International Journal of Computing
A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defec... more A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defects characterization in products. This challenging operation is very important since it is directly linked with the produced optical component’s quality. A classification phase is mandatory to complete optical devices diagnosis since a number of correctable defects are usually present beside the potential “abiding” ones. Unfortunately relevant data extracted from raw image during defects detection phase are high dimensional. This can have harmful effect on the behaviors of artificial neural networks which are suitable to perform such a challenging classification. Reducing data dimension to a smaller value can decrease the problems related to high dimensionality. In this paper we compare different techniques which permit dimensionality reduction and evaluate their impact on classification tasks performances.
Introduction The major part of the resistive forces - between 80 % and 90 % - applied to cyclist ... more Introduction The major part of the resistive forces - between 80 % and 90 % - applied to cyclist are due to the aerodynamic drag. One of the key challenge in cycling performance lies in reducing the effective frontal Area (ACd) and thus the aerodynamic drag. Therefore, ACd needs to be adequately quantified. In this regard, different methods have been proposed [1]: wind tunnel, dynamometric measurement, deceleration, and linear regression. Recently, a new approach that couples 3D digitization and computational has been investigated [1-3]. 3D model and CFD tools interest: The approach based on 3D modeling and CFD has number of advantages. First, the operating and equipment costs of this system are lower than the one of wind tunnel or linear regression. Moreover, the measuring conditions are closer to real-world testing than the other approaches. Finally, extensive experiments can be performed with only one set of digitized data, including: (I) simulating different wind and cyclist s...
Proceedings of the 2nd International Workshop on Artificial Neural Networks and Intelligent Information Processing
A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defec... more A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defects detection and characterization in products. This challenging operation is very important since it is directly linked with the produced optical component's quality. A new scratches and digs defects detection and characterization method exploiting Nomarski microscopy issued imaging has been developed. The items detected using this high-performance approach can correspond to real defects on the structure but some dusts and cleaning marks are detected too. Thus, a classification phase is necessary to complete optical devices diagnosis. In this paper, we describe a data extraction method, which supplies pertinent features from raw Nomarski images issued from industrial process. Then we apply this method to construct a database from real images. Finally we analyse the pertinence of features and the complexity of obtained database by clustering operation using an unsupervised Self Organizing Maps technique.
A major step for high-quality optical devices faults diagnosis concerns scratches and digs defect... more A major step for high-quality optical devices faults diagnosis concerns scratches and digs defects detection and characterization in products. These kinds of aesthetic flaws, shaped during different manufacturing steps, could provoke harmful effects on optical devices’ functional specificities, as well as on their optical performances by generating undesirable scatter light, which could seriously damage the expected optical features. A reliable diagnosis of these defects becomes therefore a crucial task to ensure products’ nominal specification. Moreover, such diagnosis is strongly motivated by manufacturing process correction requirements in order to guarantee mass production quality with the aim of maintaining acceptable production yield. Unfortunately, detecting and measuring such defects is still a challenging problem in production conditions and the few available automatic control solutions remain ineffective. That’s why, in most of cases, the diagnosis is performed on the basi...
It is interesting to notice that from “problem’s formulation” point of view “Industrial Computer ... more It is interesting to notice that from “problem’s formulation” point of view “Industrial Computer Aided Diagnosis” and “Biomedical Computer Aided Diagnosis” could be formulated as a same diagnosis riddle: “How point out a correct diagnosis from a set of symptoms?”. The only difference between the two above-mentioned groups of problems is the nature of the monitored (diagnosed) system: in the first group the monitored system is an artificial machinery (plant, industrial process, etc...), while in the second, the monitored system is a living body (animal or human).One of the most appealing classes of approaches allowing handling the Computer Aided Diagnosis Systems’ design in the frame of the aforementioned dual point of view is Soft-Computing based techniques, especially those dealing with neural networks and fuzzy logic. In this article, we present two soft-computing based approaches dealing with CADS design. One aims designing a biomedical oriented CADS and the other sets sights on ...
Dans differents domaines industriels, la problematique du diagnostic prend une place importante. ... more Dans differents domaines industriels, la problematique du diagnostic prend une place importante. Ainsi, le controle d’aspect des composants optiques est une etape incontournable pour garantir leurs performances operationnelles. La methode conventionnelle de controle par un operateur humain souffre de limitations importantes qui deviennent insurmontables pour certaines optiques hautes performances. Dans ce contexte, cette these traite de la conception d’un systeme automatique capable d’assurer le controle d’aspect. Premierement, une etude des capteurs pouvant etre mis en oeuvre par ce systeme est menee. Afin de satisfaire a des contraintes de temps de controle, la solution proposee utilise deux capteurs travaillant a des echelles differentes. Un de ces capteurs est base sur la microscopie Nomarski ; nous presentons ce capteur ainsi qu’un ensemble de methodes de traitement de l’image qui permettent, a partir des donnees fournies par celui-ci, de detecter les defauts et de determiner l...
Journal of Science and Cycling, 2018
Context: When riding on a flat road at high speed, the major part of the resistive forces experie... more Context: When riding on a flat road at high speed, the major part of the resistive forces experienced by a cyclist is due to aerodynamic drag. Reducing this drag is thus a key challenge for improving cycling performance. To do so it must be adequately quantified. In this regard, different methods have been proposed in literature: wind tunnel [1], dynamometric measurement [2], deceleration [3], and linear regression [4]. Recently, a new approach that couples 3D digitization and computational has been investigated [5-8].
Journal of Science and Cycling, 2019
We have previously introduced in Science Cycling 2017 an inexpensive method to assess the aerodyn... more We have previously introduced in Science Cycling 2017 an inexpensive method to assess the aerodynamic drag of a cyclist [1]. Last year we presented in the same conference a first evaluation of the method [2] where the regression method [3] was used as ground truth. In this work, the few data available had been obtained on a open road, which could increase the source of noise (wind-direction) and imprecision. In this paper we propose a consistent comparison of our method with a larger dataset. This new ground truth is still obtained with the regression method but the data were grabbed on a 200m indoor track (Bourges, France). The dataset contains records for 4 cyclists for 3 positions (top bar position, brakes hood position, bottom bar positions) at 4 different speeds (25, 30, 35, 40 km/h). All variables needed to compute aerodynamic forces are directly measured by sensors (temperature and pressure : Bosch BME280; power : Rotor inPower Powermeter; speed : Garmin 010-12103-00 speed se...
Motion capture for gesture and performance analysis in sports usually requires tracking of severa... more Motion capture for gesture and performance analysis in sports usually requires tracking of several anatomical landmarks with high spatial accuracy and sampling frequency. In addition, a model of the outer surface of the body can also be useful, e.g. for aerodynamic studies. In this context, we developed a motion capture system based on four inexpensive commercial RGB-D cameras. Our system produces both skeletal poses and 3D meshes of moving human bodies at high frequency with good accuracy.
Proceedings of the 9th International Work Conference on Artificial Neural Networks, 2007
A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defec... more A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defects characterisation. This challenging operation is very important since it is directly linked with the produced optical component’s quality. In order to automate this repetitive and difficult task, microscopy based inspection system is aimed. After a defects detection phase, a classification phase is mandatory to complete optical
Fault diagnosis of optical devices in industrial environment is a challenging but crucial task, s... more Fault diagnosis of optical devices in industrial environment is a challenging but crucial task, since it ensures products’ nominal specification and manufacturing control. Defects detection and issued information processing are among chief phases for succeeding in such diagnosis. A new scratches and digs defects detection and characterization method exploiting Nomarski microscopy issued imaging has been developed. It allows automatic check of optical devices during industrial process. Issued images contain several items which have to be detected and then classified in order to discriminate between “false” defects and “abiding” ones. In this paper, a processing method is proposed for a first step of pattern recognition from Nomarski images. A first phase permits to extract items images and a second phase allows us to cluster them using an unsupervised neural network technique, Self-Organizing Map.
A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defec... more A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defects characterisation. This challenging operation is very important since it is directly linked with the produced optical component’s quality. In order to automate this repetitive and difficult task, microscopy based inspection system is aimed. After a defects detection phase, a classification phase is mandatory to complete optical devices diagnosis because a number of correctable defects are usually present beside the potential “abiding” ones. In this paper is proposed a processing sequence, which permits to extract pertinent low-dimensional defects features from raw microscopy issued image. The described approach is validated by studying MLP neural network based classification on real industrial data using obtained defects features.
Abstract A major step for high-quality optical devices faults diagnosis concerns scratches and di... more Abstract A major step for high-quality optical devices faults diagnosis concerns scratches and digs defects detection and characterization in products. These kinds of aesthetic flaws, shaped during different manufacturing steps, could provoke harmful effects on optical ...
In this paper we investigate into perceptual properties of Style-WaveGAN, a drum synthesis method... more In this paper we investigate into perceptual properties of Style-WaveGAN, a drum synthesis method proposed in a previous publication. For both, the sound quality as well as the control precision StyleWaveGAN has been shown to deliver state of the art performance for quantitative metrics (FAD and MSE of the control parameters). The present paper aims to provide insight into the perceptual relevance of these results. Accordingly, we performed a subjective evaluation of the sound quality as well as a subjective evaluation of the precision of the control using timbre descriptors from the AudioCommons toolbox. We evaluate the sound quality with mean opinion score and make measurements of psychophysical response to the variations of the control. By means of the perceptual tests, we demonstrate that StyleWaveGAN produces better sound quality than state-of-the-art model DrumGAN and that the mean control error is lower than the absolute threshold of perception at every point of measurement used in the experiment.
Dans différents domaines industriels, la problématique du diagnostic prend une place importante. ... more Dans différents domaines industriels, la problématique du diagnostic prend une place importante. Ainsi, le contrôle d’aspect des composants optiques est une étape incontournable pour garantir leurs performances opérationnelles. La méthode conventionnelle de contrôle par un opérateur humain souffre de limitations importantes qui deviennent insurmontables pour certaines optiques hautes performances. Dans ce contexte, cette thèse traite de la conception d’un système automatique capable d’assurer le contrôle d’aspect. Premièrement, une étude des capteurs pouvant être mis en oeuvre par ce système est menée. Afin de satisfaire à des contraintes de temps de contrôle, la solution proposée utilise deux capteurs travaillant à des échelles différentes. Un de ces capteurs est basé sur la microscopie Nomarski ; nous présentons ce capteur ainsi qu’un ensemble de méthodes de traitement de l’image qui permettent, à partir des données fournies par celui-ci, de détecter les défauts et de déterminer l...
International Journal of Computing
A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defec... more A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defects characterization in products. This challenging operation is very important since it is directly linked with the produced optical component’s quality. A classification phase is mandatory to complete optical devices diagnosis since a number of correctable defects are usually present beside the potential “abiding” ones. Unfortunately relevant data extracted from raw image during defects detection phase are high dimensional. This can have harmful effect on the behaviors of artificial neural networks which are suitable to perform such a challenging classification. Reducing data dimension to a smaller value can decrease the problems related to high dimensionality. In this paper we compare different techniques which permit dimensionality reduction and evaluate their impact on classification tasks performances.
Introduction The major part of the resistive forces - between 80 % and 90 % - applied to cyclist ... more Introduction The major part of the resistive forces - between 80 % and 90 % - applied to cyclist are due to the aerodynamic drag. One of the key challenge in cycling performance lies in reducing the effective frontal Area (ACd) and thus the aerodynamic drag. Therefore, ACd needs to be adequately quantified. In this regard, different methods have been proposed [1]: wind tunnel, dynamometric measurement, deceleration, and linear regression. Recently, a new approach that couples 3D digitization and computational has been investigated [1-3]. 3D model and CFD tools interest: The approach based on 3D modeling and CFD has number of advantages. First, the operating and equipment costs of this system are lower than the one of wind tunnel or linear regression. Moreover, the measuring conditions are closer to real-world testing than the other approaches. Finally, extensive experiments can be performed with only one set of digitized data, including: (I) simulating different wind and cyclist s...
Proceedings of the 2nd International Workshop on Artificial Neural Networks and Intelligent Information Processing
A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defec... more A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defects detection and characterization in products. This challenging operation is very important since it is directly linked with the produced optical component's quality. A new scratches and digs defects detection and characterization method exploiting Nomarski microscopy issued imaging has been developed. The items detected using this high-performance approach can correspond to real defects on the structure but some dusts and cleaning marks are detected too. Thus, a classification phase is necessary to complete optical devices diagnosis. In this paper, we describe a data extraction method, which supplies pertinent features from raw Nomarski images issued from industrial process. Then we apply this method to construct a database from real images. Finally we analyse the pertinence of features and the complexity of obtained database by clustering operation using an unsupervised Self Organizing Maps technique.
A major step for high-quality optical devices faults diagnosis concerns scratches and digs defect... more A major step for high-quality optical devices faults diagnosis concerns scratches and digs defects detection and characterization in products. These kinds of aesthetic flaws, shaped during different manufacturing steps, could provoke harmful effects on optical devices’ functional specificities, as well as on their optical performances by generating undesirable scatter light, which could seriously damage the expected optical features. A reliable diagnosis of these defects becomes therefore a crucial task to ensure products’ nominal specification. Moreover, such diagnosis is strongly motivated by manufacturing process correction requirements in order to guarantee mass production quality with the aim of maintaining acceptable production yield. Unfortunately, detecting and measuring such defects is still a challenging problem in production conditions and the few available automatic control solutions remain ineffective. That’s why, in most of cases, the diagnosis is performed on the basi...
It is interesting to notice that from “problem’s formulation” point of view “Industrial Computer ... more It is interesting to notice that from “problem’s formulation” point of view “Industrial Computer Aided Diagnosis” and “Biomedical Computer Aided Diagnosis” could be formulated as a same diagnosis riddle: “How point out a correct diagnosis from a set of symptoms?”. The only difference between the two above-mentioned groups of problems is the nature of the monitored (diagnosed) system: in the first group the monitored system is an artificial machinery (plant, industrial process, etc...), while in the second, the monitored system is a living body (animal or human).One of the most appealing classes of approaches allowing handling the Computer Aided Diagnosis Systems’ design in the frame of the aforementioned dual point of view is Soft-Computing based techniques, especially those dealing with neural networks and fuzzy logic. In this article, we present two soft-computing based approaches dealing with CADS design. One aims designing a biomedical oriented CADS and the other sets sights on ...
Dans differents domaines industriels, la problematique du diagnostic prend une place importante. ... more Dans differents domaines industriels, la problematique du diagnostic prend une place importante. Ainsi, le controle d’aspect des composants optiques est une etape incontournable pour garantir leurs performances operationnelles. La methode conventionnelle de controle par un operateur humain souffre de limitations importantes qui deviennent insurmontables pour certaines optiques hautes performances. Dans ce contexte, cette these traite de la conception d’un systeme automatique capable d’assurer le controle d’aspect. Premierement, une etude des capteurs pouvant etre mis en oeuvre par ce systeme est menee. Afin de satisfaire a des contraintes de temps de controle, la solution proposee utilise deux capteurs travaillant a des echelles differentes. Un de ces capteurs est base sur la microscopie Nomarski ; nous presentons ce capteur ainsi qu’un ensemble de methodes de traitement de l’image qui permettent, a partir des donnees fournies par celui-ci, de detecter les defauts et de determiner l...
Journal of Science and Cycling, 2018
Context: When riding on a flat road at high speed, the major part of the resistive forces experie... more Context: When riding on a flat road at high speed, the major part of the resistive forces experienced by a cyclist is due to aerodynamic drag. Reducing this drag is thus a key challenge for improving cycling performance. To do so it must be adequately quantified. In this regard, different methods have been proposed in literature: wind tunnel [1], dynamometric measurement [2], deceleration [3], and linear regression [4]. Recently, a new approach that couples 3D digitization and computational has been investigated [5-8].
Journal of Science and Cycling, 2019
We have previously introduced in Science Cycling 2017 an inexpensive method to assess the aerodyn... more We have previously introduced in Science Cycling 2017 an inexpensive method to assess the aerodynamic drag of a cyclist [1]. Last year we presented in the same conference a first evaluation of the method [2] where the regression method [3] was used as ground truth. In this work, the few data available had been obtained on a open road, which could increase the source of noise (wind-direction) and imprecision. In this paper we propose a consistent comparison of our method with a larger dataset. This new ground truth is still obtained with the regression method but the data were grabbed on a 200m indoor track (Bourges, France). The dataset contains records for 4 cyclists for 3 positions (top bar position, brakes hood position, bottom bar positions) at 4 different speeds (25, 30, 35, 40 km/h). All variables needed to compute aerodynamic forces are directly measured by sensors (temperature and pressure : Bosch BME280; power : Rotor inPower Powermeter; speed : Garmin 010-12103-00 speed se...
Motion capture for gesture and performance analysis in sports usually requires tracking of severa... more Motion capture for gesture and performance analysis in sports usually requires tracking of several anatomical landmarks with high spatial accuracy and sampling frequency. In addition, a model of the outer surface of the body can also be useful, e.g. for aerodynamic studies. In this context, we developed a motion capture system based on four inexpensive commercial RGB-D cameras. Our system produces both skeletal poses and 3D meshes of moving human bodies at high frequency with good accuracy.
Proceedings of the 9th International Work Conference on Artificial Neural Networks, 2007
A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defec... more A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defects characterisation. This challenging operation is very important since it is directly linked with the produced optical component’s quality. In order to automate this repetitive and difficult task, microscopy based inspection system is aimed. After a defects detection phase, a classification phase is mandatory to complete optical
Fault diagnosis of optical devices in industrial environment is a challenging but crucial task, s... more Fault diagnosis of optical devices in industrial environment is a challenging but crucial task, since it ensures products’ nominal specification and manufacturing control. Defects detection and issued information processing are among chief phases for succeeding in such diagnosis. A new scratches and digs defects detection and characterization method exploiting Nomarski microscopy issued imaging has been developed. It allows automatic check of optical devices during industrial process. Issued images contain several items which have to be detected and then classified in order to discriminate between “false” defects and “abiding” ones. In this paper, a processing method is proposed for a first step of pattern recognition from Nomarski images. A first phase permits to extract items images and a second phase allows us to cluster them using an unsupervised neural network technique, Self-Organizing Map.
A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defec... more A major step for high-quality optical surfaces faults diagnosis concerns scratches and digs defects characterisation. This challenging operation is very important since it is directly linked with the produced optical component’s quality. In order to automate this repetitive and difficult task, microscopy based inspection system is aimed. After a defects detection phase, a classification phase is mandatory to complete optical devices diagnosis because a number of correctable defects are usually present beside the potential “abiding” ones. In this paper is proposed a processing sequence, which permits to extract pertinent low-dimensional defects features from raw microscopy issued image. The described approach is validated by studying MLP neural network based classification on real industrial data using obtained defects features.
Abstract A major step for high-quality optical devices faults diagnosis concerns scratches and di... more Abstract A major step for high-quality optical devices faults diagnosis concerns scratches and digs defects detection and characterization in products. These kinds of aesthetic flaws, shaped during different manufacturing steps, could provoke harmful effects on optical ...