The automatic extraction of the notes that were played in a digital musical signal (automatic music transcription) is an open problem. A number of techniques have been applied to solve it without concluding results. T...
详细信息
The automatic extraction of the notes that were played in a digital musical signal (automatic music transcription) is an open problem. A number of techniques have been applied to solve it without concluding results. The monotimbral polyphonic version of the problem is posed here: a single instrument has been played and more than one note can sound at the same time. This work tries to approach it through the identification of the pattern of a given instrument in the frequency domain. This is achieved using time-delay neuralnetworks that are fed with the band-grouped spectrogram of a polyphonic monotimbral music recording. The use of a learning scheme based on examples like neuralnetworks permits our system to avoid the use of an auditory model to approach this problem. A number of issues have to be faced to have a robust and powerful system, but promising results using synthesized instruments are presented. (c) 2005 Elsevier B.V. All rights reserved.
Localizing faces in images is a difficult task, and represents the firststep towards the solution of the face recognition problem. Moreover, devising an effective face detection method can provide some suggestions to...
详细信息
Localizing faces in images is a difficult task, and represents the firststep towards the solution of the face recognition problem. Moreover, devising an effective face detection method can provide some suggestions to solve similar object and pattern detection problems. This paper presents a novel approach to the solution of the face localization problem using Recursive neuralnetworks (RNNs). The proposed method assumes a graph-based representation of images that combines structural and symbolic visual features. Such graphs are then processed by RNNs, in order to establish the possible presence and the position of faces inside the image. A novel RNN model that can deal with graphs with labeled edges has been also exploited. Some experiments on snapshots from video sequences are reported, showing very promising results. (c) 2005 Elsevier B.V. All rights reserved.
This paper introduces a new class of sign-based training algorithms for neuralnetworks that combine the sign-based updates of the Rprop algorithm with the composite nonlinear Jacobi method. The theoretical foundation...
详细信息
This paper introduces a new class of sign-based training algorithms for neuralnetworks that combine the sign-based updates of the Rprop algorithm with the composite nonlinear Jacobi method. The theoretical foundations of the class are described and a heuristic Rprop-based Jacobi algorithm is empirically investigated through simulation experiments in benchmark pattern classification problems. Numerical evidence shows that this new modification of the Rprop algorithm exhibits improved learning speed in all cases tested, and compares favorably against the Rprop and a recently proposed modification, the improved Rprop. (c) 2005 Elsevier B.V. All rights reserved.
in semiconductor manufacturing, the spatial pattern of failed devices in a wafer can give precious hints on which step of the process is responsible for the failures. In the literature, Kohonen39;s Self Organizing F...
详细信息
in semiconductor manufacturing, the spatial pattern of failed devices in a wafer can give precious hints on which step of the process is responsible for the failures. In the literature, Kohonen's Self Organizing Feature Maps (SOM) and Adaptive Resonance Theory 1 (ART1) architectures have been compared, concluding that the latter are to be preferred. However, both the simulated and the real data sets used for validation and comparison were very limited. In this paper, the use of ART1 and SOM as wafer classifiers is re-assessed on much more extensive simulated and real data sets. We conclude that ART1 is not adequate, whereas SOM provide completely satisfactory results including visually effective representation of spatial failure probability of the pattern classes. (c) 2005 Elsevier B.V. All rights reserved.
Feature selection is a fundamental process in many classifier design problems. However, it is NP-complete and approximate approaches often require requires extensive exploration and evaluation. This paper describes a ...
详细信息
Feature selection is a fundamental process in many classifier design problems. However, it is NP-complete and approximate approaches often require requires extensive exploration and evaluation. This paper describes a novel approach that represents feature selection as a continuous regularization problem which has a single, global minimum, where the model's complexity is measured using a 1-norm on the parameter vector. A new exploratory design process is also described that allows the designer to efficiently construct the complete locus of sparse, kernel-based classifiers. It allows the designer to investigate the optimal parameters' trajectories as the regularization parameter is altered and look for effects, such as Simpson's paradox, that occur in many multivariate data analysis problems. The approach is demonstrated on the well-known Australian Credit data set. (c) 2005 Published by Elsevier B.V.
Convolution kernels and recursive neuralnetworks are both suitable approaches for supervised learning when the input is a discrete structure like a labeled tree or graph. We compare these techniques in two natural la...
详细信息
Convolution kernels and recursive neuralnetworks are both suitable approaches for supervised learning when the input is a discrete structure like a labeled tree or graph. We compare these techniques in two natural language problems. In both problems, the learning task consists in choosing the best alternative tree in a set of candidates. We report about an empirical evaluation between the two methods on a large corpus of parsed sentences and speculate on the role played by the representation and the loss function. (c) 2005 Elsevier B.V. All rights reserved.
Kohonen self-organisation maps are a well know classification tool, commonly used in a wide variety of problems, but with limited applications in time series forecasting context. In this paper, we propose a forecastin...
详细信息
Kohonen self-organisation maps are a well know classification tool, commonly used in a wide variety of problems, but with limited applications in time series forecasting context. In this paper, we propose a forecasting method specifically designed for multi-dimensional long-term trends prediction, with a double application of the Kohonen algorithm. Practical applications of the method are also presented. (c) 2005 Elsevier B.V. All rights reserved.
This paper describes several techniques improving a Chinese character recognition system. Enhanced nonlinear normalization, feature extraction and tuning kernel parameters of support vector machine on a large data set...
详细信息
This paper describes several techniques improving a Chinese character recognition system. Enhanced nonlinear normalization, feature extraction and tuning kernel parameters of support vector machine on a large data set with thousands of classes, contribute to improvement of the overall system performance. The enhanced nonlinear normalization method not only solves the aliasing problem in the original Yamada et al.'s nonlinear normalization method but also avoids the undue stroke distortion in the peripheral region of the normalized image. The support vector machine is for the first time tested on a large data set composed of several million samples and thousands of classes. The recognition system has achieved a high recognition rate of 99.0% on ETL9B, a handwritten Chinese character database. (c) 2005 Elsevier B.V. All rights reserved.
Handwriting recognition for hand-held devices like PDAs requires very accurate and adaptive classifiers. It is such a complex classification problem that it is quite usual now to make co-operate several classification...
详细信息
Handwriting recognition for hand-held devices like PDAs requires very accurate and adaptive classifiers. It is such a complex classification problem that it is quite usual now to make co-operate several classification methods. In this paper, we present an original two stages recognizer. The firststage is a model-based classifier which store an exhaustive set of character models. The second stage is a pairwise classifier which separate the most ambiguous pairs of classes. This hybrid architecture is based on the idea that the correct class almost systematically belongs to the two more relevant classes found by the first classifier. Experiments on a 80,000 examples database show a 30% improvement on a 62 classes recognition problem. Moreover, we show experimentally that such an architecture suits perfectly for incremental classification. (c) 2005 Elsevier B.V. All rights reserved.
Feature selection is a fundamental process in many classifier design problems. However, it is NP-complete and approximate approaches often require requires extensive exploration and evaluation. This paper describes a ...
详细信息
Feature selection is a fundamental process in many classifier design problems. However, it is NP-complete and approximate approaches often require requires extensive exploration and evaluation. This paper describes a novel approach that represents feature selection as a continuous regularization problem which has a single, global minimum, where the model's complexity is measured using a 1-norm on the parameter vector. A new exploratory design process is also described that allows the designer to efficiently construct the complete locus of sparse, kernel-based classifiers. It allows the designer to investigate the optimal parameters' trajectories as the regularization parameter is altered and look for effects, such as Simpson's paradox, that occur in many multivariate data analysis problems. The approach is demonstrated on the well-known Australian Credit data set. (c) 2005 Published by Elsevier B.V.
暂无评论