Jose Iñesta - Academia.edu (original) (raw)
Papers by Jose Iñesta
The success of the Internet has filled the net with lots of symbolic representations of music wor... more The success of the Internet has filled the net with lots of symbolic representations of music works. Two kinds of problems arise to the user: content-based search of music and the identification of similar works. Both belong to the pattern recognition domain. In contrast to most of the existing approaches, we pose a non-linear representation of a melody, based on trees that express the metric and rhythm of music in a natural way. This representation provide a number of advantages: more musical significance, more compact representation and others. Here we have worked on the comparison of melodies for identification.
The success of the Internet has filled the net with lots of symbolic representations of music wor... more The success of the Internet has filled the net with lots of symbolic representations of music works. Two kinds of problems arise to the user: content-based search of music and the identification of similar works. Both belong to the pattern recognition domain. In contrast to most of the existing approaches, we pose a non-linear representation of a melody, based on trees that express the metric and rhythm of music in a natural way. This representation provide a number of advantages: more musical significance, more compact representation and others. Here we have worked on the comparison of melodies for identification.
Evolutionary methods have been largely used in algorithmic music composition due to their ability... more Evolutionary methods have been largely used in algorithmic music composition due to their ability to explore an immense space of possibilities. The main problem of genetic related composition algorithms has always been the implementation of the selection process. In this work, a pattern recognition-based system helped by a number of music analysis rules is designed for that task. The fitness value provided by this kind of supervisor (the music "critic") models the affect for a certain music genre after a training phase. The early stages of this work have been encouraging since they have responded to the a priori expectations and more work has to be carried out in the future to explore the creative capabilities of the proposed system.
Melodic similarity is an important research topic in music information retrieval. The representat... more Melodic similarity is an important research topic in music information retrieval. The representation of symbolic music by means of trees has proven to be suitable in melodic similarity computation, because they are able to code rhythm in their structure leaving only pitch representations as a degree of freedom for coding. In order to compare trees, different edit distances have been previously used. In this paper, stochastic k-testable tree-models, formerly used in other domains like structured document compression or natural language processing, have been used for computing a similarity measure between melody trees as a probability and their performance has been compared to a classical tree edit distance.
The success of the Internet has filled the net with lots of symbolic representations of music wor... more The success of the Internet has filled the net with lots of symbolic representations of music works. Two kinds of problems arise to the user: the search for information from content and the identification of similar works. Both belong to the pattern recognition domain. In contrast to most of the existing approaches, we pose a non-linear representation of a melody, based on trees that express the metric and rhythm of music in a natural way. This representation provide a number of advantages: more musical significance, more compact representation and others. Here we have worked on the comparison of melodies and patterns, leading to motive extraction and its use for the identification of complete melodies.
The success of the Internet has filled the net with lots of symbolic representations of music wor... more The success of the Internet has filled the net with lots of symbolic representations of music works. Two kinds of problems arise to the user: the search for information from content and the identification of similar works. Both belong to the pattern recognition domain. In contrast to most of the existing approaches, we pose a non-linear representation of a melody, based on trees that express the metric and rhythm of music in a natural way. This representation provide a number of advantages: more musical significance, more compact representation and others. Here we have worked on the comparison of melodies and patterns, leading to motive extraction and its use for the identification of complete melodies.
The tree representation, using rhythm for defining the tree structure and pitch infor-mation for ... more The tree representation, using rhythm for defining the tree structure and pitch infor-mation for node labeling has proven to be ef-fective in melodic similarity computation. In this paper we propose a solution representing melodies by tree grammars. For that, we in-fer a probabilistic context-free grammars for the melodies in a database, using their tree coding (with duration and pitch) and classify queries represented as a string of pitches. We aim to assess their ability to identify a noisy snippet query among a set of songs stored in symbolic format.
Through a 3D reconstruction of the human back surface using structured light techniques, we study... more Through a 3D reconstruction of the human back surface using structured light techniques, we study the properties of spine curve by means of a set of parameters related to measures commonly applied in medicine. In this way, descriptors for measuring the abnormalities in the projections of the front and sagittal planes can be computed. We build the spine curve in 3D and analyse the behaviour of the Frenet frame when along the curve the deformation processes in idiophatic scoliosis appear.
A novel approach is presented to computing all possible grips with "contact stability" ... more A novel approach is presented to computing all possible grips with "contact stability" on unknown objects by means of visual parameters as input. This approach extends a previous one that obtains grips under more restrictive stability conditions. The complete application of all this grasping information is out of the scope of this paper, because requires a high level decision by means of a task-planner with knowledge of a specific task in a certain context. An end-effector mounted camera provides the robot with 2D images of the object so that it can manipulate the object using a parallel-jaw gripper. The approach has been implemented, and experimental results are provided showing the reliability and the required computational effort.
We present preliminary work on automatic human-readable melody characterization. In order to obta... more We present preliminary work on automatic human-readable melody characterization. In order to obtain such a characterization, we (1) extract a set of statistical descriptors from the tracks in a dataset of MIDI files, (2) apply a rule induction algorithm to obtain a set of (crisp) classification rules for melody track identification, and (3) automatically transform the crisp rules into fuzzy rules by applying a genetic algorithm to generate the membership functions for the rule attributes. Some results are presented and discussed. This work is supported by the projects: GV06/166 and CICyT TIN2006–14932–C02, partially supported by EU ERDF.
Previous work in genre recognition and characterization from symbolic sources (melodies extracted... more Previous work in genre recognition and characterization from symbolic sources (melodies extracted from MIDI files) car-ried out by our group pointed our research to study how the different utilized approaches perform and how their differ-ent abilities can be used together in order to improve both the accuracy and robustness of their decisions. Results for a corpus of Jazz and Classical music pieces are presented and discussed.
Previous work done in genre recognition and characterization from sym- bolic sources (monophonic ... more Previous work done in genre recognition and characterization from sym- bolic sources (monophonic melodies extracted from MIDI files) have pointed our research to the use of classifier ensembles to better accomplish the task. This work presents current research in the use of voting ensembles of classifiers trained on statistical description models of melodies, in order to improve both the accuracy and robustness of single classifier systems in the genre recognition task. Different voting schemes are discussed and compared, and results for a corpus of Jazz and Classical music pieces are presented and assesed. Keywords: Statistical pattern recognition, Classifier ensembles, Music infor- mation retrieval, Musical genre recognition
We present a cartesian ensemble classification system that is based on the principle of late fusi... more We present a cartesian ensemble classification system that is based on the principle of late fusion and feature subspaces. These feature subspaces describe different aspects of the same data set. The framework is built on the Weka machine learning toolkit and able to combine arbitrary feature sets and learning schemes. In our scenario, we use it for the ensemble classification of multiple feature sets from the audio and symbolic domains. We present an extensive set of experiments in the context of music genre classification, based on numerous Music IR benchmark datasets, and evaluate a set of combination/voting rules. The results show that the approach is superior to the best choice of a single algorithm on a single feature set. Moreover, it also releases the user from making this choice explicitly.
... This work tries to generalise our former results obtained with a single DPD algorithm and loo... more ... This work tries to generalise our former results obtained with a single DPD algorithm and look for a network architecture able to provide the best results regardless of the algorithm considered. ... Table 5. Best resulting topologies from the study for each algorithm. ...
Lecture Notes in Computer Science, 1998
Although telerobotic systems are becoming more complex, there are few actions they can perform on... more Although telerobotic systems are becoming more complex, there are few actions they can perform on their own and, moreover, knowledge about the tasks they are being used for often relies only on their operator. In this paper, we present the design of a telerobotic system that features learning capabilities, can accept commands given in natural language and provides control of
Proceedings of 3rd international workshop on Machine learning and music - MML '10, 2010
We evaluate the impact of feature selection on the classification accuracy and the achieved dimen... more We evaluate the impact of feature selection on the classification accuracy and the achieved dimensionality reduction, which benefits the time needed on training classification models. Our classification scheme therein is a Cartesian ensemble classification system, based on the principle of late fusion and feature subspaces. These feature subspaces describe different aspects of the same data set. We use it for the ensemble classification of multiple feature sets from the audio and symbolic domains. We present an extensive set of experiments in the context of music genre classification, based on Music IR benchmark datasets. We show that while feature selection does not benefit classification accuracy, it greatly reduces the dimensionality of each feature subspace, and thus adds to great gains in the time needed to train the individual classification models that form the ensemble.
Lecture Notes in Computer Science, 2005
The automatic classification of music files into styles is one challenging problem in music infor... more The automatic classification of music files into styles is one challenging problem in music information retrieval and for music style perception understanding. It has a number of applications, like the indexation and exploration of musical databases. Some techniques used in text classification can be applied to this problem. The key point is to establish a music equivalent to the words in texts. A number of works use the combination of intervals and duration ratios for music description. In this paper, different statistical text recognition algorithms are applied to style recognition using this kind of melody representation, exploring their performance for different word sizes and statistical models.
Journal of New Music Research, 2014
The success of the Internet has filled the net with lots of symbolic representations of music wor... more The success of the Internet has filled the net with lots of symbolic representations of music works. Two kinds of problems arise to the user: content-based search of music and the identification of similar works. Both belong to the pattern recognition domain. In contrast to most of the existing approaches, we pose a non-linear representation of a melody, based on trees that express the metric and rhythm of music in a natural way. This representation provide a number of advantages: more musical significance, more compact representation and others. Here we have worked on the comparison of melodies for identification.
The success of the Internet has filled the net with lots of symbolic representations of music wor... more The success of the Internet has filled the net with lots of symbolic representations of music works. Two kinds of problems arise to the user: content-based search of music and the identification of similar works. Both belong to the pattern recognition domain. In contrast to most of the existing approaches, we pose a non-linear representation of a melody, based on trees that express the metric and rhythm of music in a natural way. This representation provide a number of advantages: more musical significance, more compact representation and others. Here we have worked on the comparison of melodies for identification.
Evolutionary methods have been largely used in algorithmic music composition due to their ability... more Evolutionary methods have been largely used in algorithmic music composition due to their ability to explore an immense space of possibilities. The main problem of genetic related composition algorithms has always been the implementation of the selection process. In this work, a pattern recognition-based system helped by a number of music analysis rules is designed for that task. The fitness value provided by this kind of supervisor (the music "critic") models the affect for a certain music genre after a training phase. The early stages of this work have been encouraging since they have responded to the a priori expectations and more work has to be carried out in the future to explore the creative capabilities of the proposed system.
Melodic similarity is an important research topic in music information retrieval. The representat... more Melodic similarity is an important research topic in music information retrieval. The representation of symbolic music by means of trees has proven to be suitable in melodic similarity computation, because they are able to code rhythm in their structure leaving only pitch representations as a degree of freedom for coding. In order to compare trees, different edit distances have been previously used. In this paper, stochastic k-testable tree-models, formerly used in other domains like structured document compression or natural language processing, have been used for computing a similarity measure between melody trees as a probability and their performance has been compared to a classical tree edit distance.
The success of the Internet has filled the net with lots of symbolic representations of music wor... more The success of the Internet has filled the net with lots of symbolic representations of music works. Two kinds of problems arise to the user: the search for information from content and the identification of similar works. Both belong to the pattern recognition domain. In contrast to most of the existing approaches, we pose a non-linear representation of a melody, based on trees that express the metric and rhythm of music in a natural way. This representation provide a number of advantages: more musical significance, more compact representation and others. Here we have worked on the comparison of melodies and patterns, leading to motive extraction and its use for the identification of complete melodies.
The success of the Internet has filled the net with lots of symbolic representations of music wor... more The success of the Internet has filled the net with lots of symbolic representations of music works. Two kinds of problems arise to the user: the search for information from content and the identification of similar works. Both belong to the pattern recognition domain. In contrast to most of the existing approaches, we pose a non-linear representation of a melody, based on trees that express the metric and rhythm of music in a natural way. This representation provide a number of advantages: more musical significance, more compact representation and others. Here we have worked on the comparison of melodies and patterns, leading to motive extraction and its use for the identification of complete melodies.
The tree representation, using rhythm for defining the tree structure and pitch infor-mation for ... more The tree representation, using rhythm for defining the tree structure and pitch infor-mation for node labeling has proven to be ef-fective in melodic similarity computation. In this paper we propose a solution representing melodies by tree grammars. For that, we in-fer a probabilistic context-free grammars for the melodies in a database, using their tree coding (with duration and pitch) and classify queries represented as a string of pitches. We aim to assess their ability to identify a noisy snippet query among a set of songs stored in symbolic format.
Through a 3D reconstruction of the human back surface using structured light techniques, we study... more Through a 3D reconstruction of the human back surface using structured light techniques, we study the properties of spine curve by means of a set of parameters related to measures commonly applied in medicine. In this way, descriptors for measuring the abnormalities in the projections of the front and sagittal planes can be computed. We build the spine curve in 3D and analyse the behaviour of the Frenet frame when along the curve the deformation processes in idiophatic scoliosis appear.
A novel approach is presented to computing all possible grips with "contact stability" ... more A novel approach is presented to computing all possible grips with "contact stability" on unknown objects by means of visual parameters as input. This approach extends a previous one that obtains grips under more restrictive stability conditions. The complete application of all this grasping information is out of the scope of this paper, because requires a high level decision by means of a task-planner with knowledge of a specific task in a certain context. An end-effector mounted camera provides the robot with 2D images of the object so that it can manipulate the object using a parallel-jaw gripper. The approach has been implemented, and experimental results are provided showing the reliability and the required computational effort.
We present preliminary work on automatic human-readable melody characterization. In order to obta... more We present preliminary work on automatic human-readable melody characterization. In order to obtain such a characterization, we (1) extract a set of statistical descriptors from the tracks in a dataset of MIDI files, (2) apply a rule induction algorithm to obtain a set of (crisp) classification rules for melody track identification, and (3) automatically transform the crisp rules into fuzzy rules by applying a genetic algorithm to generate the membership functions for the rule attributes. Some results are presented and discussed. This work is supported by the projects: GV06/166 and CICyT TIN2006–14932–C02, partially supported by EU ERDF.
Previous work in genre recognition and characterization from symbolic sources (melodies extracted... more Previous work in genre recognition and characterization from symbolic sources (melodies extracted from MIDI files) car-ried out by our group pointed our research to study how the different utilized approaches perform and how their differ-ent abilities can be used together in order to improve both the accuracy and robustness of their decisions. Results for a corpus of Jazz and Classical music pieces are presented and discussed.
Previous work done in genre recognition and characterization from sym- bolic sources (monophonic ... more Previous work done in genre recognition and characterization from sym- bolic sources (monophonic melodies extracted from MIDI files) have pointed our research to the use of classifier ensembles to better accomplish the task. This work presents current research in the use of voting ensembles of classifiers trained on statistical description models of melodies, in order to improve both the accuracy and robustness of single classifier systems in the genre recognition task. Different voting schemes are discussed and compared, and results for a corpus of Jazz and Classical music pieces are presented and assesed. Keywords: Statistical pattern recognition, Classifier ensembles, Music infor- mation retrieval, Musical genre recognition
We present a cartesian ensemble classification system that is based on the principle of late fusi... more We present a cartesian ensemble classification system that is based on the principle of late fusion and feature subspaces. These feature subspaces describe different aspects of the same data set. The framework is built on the Weka machine learning toolkit and able to combine arbitrary feature sets and learning schemes. In our scenario, we use it for the ensemble classification of multiple feature sets from the audio and symbolic domains. We present an extensive set of experiments in the context of music genre classification, based on numerous Music IR benchmark datasets, and evaluate a set of combination/voting rules. The results show that the approach is superior to the best choice of a single algorithm on a single feature set. Moreover, it also releases the user from making this choice explicitly.
... This work tries to generalise our former results obtained with a single DPD algorithm and loo... more ... This work tries to generalise our former results obtained with a single DPD algorithm and look for a network architecture able to provide the best results regardless of the algorithm considered. ... Table 5. Best resulting topologies from the study for each algorithm. ...
Lecture Notes in Computer Science, 1998
Although telerobotic systems are becoming more complex, there are few actions they can perform on... more Although telerobotic systems are becoming more complex, there are few actions they can perform on their own and, moreover, knowledge about the tasks they are being used for often relies only on their operator. In this paper, we present the design of a telerobotic system that features learning capabilities, can accept commands given in natural language and provides control of
Proceedings of 3rd international workshop on Machine learning and music - MML '10, 2010
We evaluate the impact of feature selection on the classification accuracy and the achieved dimen... more We evaluate the impact of feature selection on the classification accuracy and the achieved dimensionality reduction, which benefits the time needed on training classification models. Our classification scheme therein is a Cartesian ensemble classification system, based on the principle of late fusion and feature subspaces. These feature subspaces describe different aspects of the same data set. We use it for the ensemble classification of multiple feature sets from the audio and symbolic domains. We present an extensive set of experiments in the context of music genre classification, based on Music IR benchmark datasets. We show that while feature selection does not benefit classification accuracy, it greatly reduces the dimensionality of each feature subspace, and thus adds to great gains in the time needed to train the individual classification models that form the ensemble.
Lecture Notes in Computer Science, 2005
The automatic classification of music files into styles is one challenging problem in music infor... more The automatic classification of music files into styles is one challenging problem in music information retrieval and for music style perception understanding. It has a number of applications, like the indexation and exploration of musical databases. Some techniques used in text classification can be applied to this problem. The key point is to establish a music equivalent to the words in texts. A number of works use the combination of intervals and duration ratios for music description. In this paper, different statistical text recognition algorithms are applied to style recognition using this kind of melody representation, exploring their performance for different word sizes and statistical models.
Journal of New Music Research, 2014