In the field of sparse representations, the overcomplete dictionary learning problem is of crucial importance and has a growing application pool where it is used. In this paper we present an iterative dictionary learning algorithm based on the singular value decomposition that efficiently construct unions of orthonormal bases. The important innovation described in this paper, that affects positively the running time of the learning procedures, is the way in which the sparse representations are computed - data are reconstructed in a single orthonormal base, avoiding slow sparse approximation algorithms - how the bases in the union are used and updated individually and how the union itself is expanded by looking at the worst reconstructed data items. The numerical experiments show conclusively the speedup induced by our method when compared to previous works, for the same target representation error.
{"title":"Block orthonormal overcomplete dictionary learning","authors":"Cristian Rusu, B. Dumitrescu","doi":"10.5281/ZENODO.43438","DOIUrl":"https://doi.org/10.5281/ZENODO.43438","url":null,"abstract":"In the field of sparse representations, the overcomplete dictionary learning problem is of crucial importance and has a growing application pool where it is used. In this paper we present an iterative dictionary learning algorithm based on the singular value decomposition that efficiently construct unions of orthonormal bases. The important innovation described in this paper, that affects positively the running time of the learning procedures, is the way in which the sparse representations are computed - data are reconstructed in a single orthonormal base, avoiding slow sparse approximation algorithms - how the bases in the union are used and updated individually and how the union itself is expanded by looking at the worst reconstructed data items. The numerical experiments show conclusively the speedup induced by our method when compared to previous works, for the same target representation error.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"146 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116487208","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
An operation planning system, integrating dynamic environmental forecasts and satellite Automatic Identification System sensor performance surfaces, to improve maritime traffic situation awareness is proposed and tested. Multi-objective evolutionary algorithms are used to optimize a network of monitoring assets with respect to a combined surveillance and piracy activity risk metric, the network area coverage and the mission cost, under given spatial and kinematic constraints. Pareto efficient solutions are provided, each representing a tradeoff among mission objectives. Tests in a counter piracy operational scenario with real-world hindcast data and sensor performance surfaces show the effectiveness of the methodology in improving surveillance efficiency.
{"title":"Asset network planning: Integration of environmental data and sensor performance for counter piracy","authors":"R. Grasso, P. Braca, J. Osler, J. Hansen","doi":"10.5281/ZENODO.43743","DOIUrl":"https://doi.org/10.5281/ZENODO.43743","url":null,"abstract":"An operation planning system, integrating dynamic environmental forecasts and satellite Automatic Identification System sensor performance surfaces, to improve maritime traffic situation awareness is proposed and tested. Multi-objective evolutionary algorithms are used to optimize a network of monitoring assets with respect to a combined surveillance and piracy activity risk metric, the network area coverage and the mission cost, under given spatial and kinematic constraints. Pareto efficient solutions are provided, each representing a tradeoff among mission objectives. Tests in a counter piracy operational scenario with real-world hindcast data and sensor performance surfaces show the effectiveness of the methodology in improving surveillance efficiency.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124491031","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
To identify the scalp projections of the underlying sources of neural activity based on recorded electroencephalographic (EEG) signals, the multi-dimensional decomposition models Parallel Factor Analysis (PARAFAC) and Parallel Factor Analysis 2 (PARAFAC2) have recently attained interest. We evaluate the models based on synthetic EEG data, because this allows an objective assessment by comparing the estimated projections to the parameters of the sources. We simulate EEG data using the EEG forward solution and focus on dynamic sources that change their spatial projection over time. Recently, this type of signal has been identified as the dominant type of signal, e. g. in measurements of visual evoked potentials. Further, we develop a method to objectively evaluate the decomposition models. We show that the decomposition models reconstruct the scalp projections successfully from data with low signal-to-noise ratio (SNR). They perform best if the number of calculated components (model order) equals the number of sources.
{"title":"Evaluation of multi-dimensional decomposition models using synthetic moving EEG potentials","authors":"J. Mengelkamp, M. Weis, P. Husar","doi":"10.5281/ZENODO.43441","DOIUrl":"https://doi.org/10.5281/ZENODO.43441","url":null,"abstract":"To identify the scalp projections of the underlying sources of neural activity based on recorded electroencephalographic (EEG) signals, the multi-dimensional decomposition models Parallel Factor Analysis (PARAFAC) and Parallel Factor Analysis 2 (PARAFAC2) have recently attained interest. We evaluate the models based on synthetic EEG data, because this allows an objective assessment by comparing the estimated projections to the parameters of the sources. We simulate EEG data using the EEG forward solution and focus on dynamic sources that change their spatial projection over time. Recently, this type of signal has been identified as the dominant type of signal, e. g. in measurements of visual evoked potentials. Further, we develop a method to objectively evaluate the decomposition models. We show that the decomposition models reconstruct the scalp projections successfully from data with low signal-to-noise ratio (SNR). They perform best if the number of calculated components (model order) equals the number of sources.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124496911","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We propose a theoretical study of the conditions guaranteeing that a decoder will obtain an optimal signal recovery from an underdetermined set of linear measurements. This special type of performance guarantee is termed instance optimality and is typically related with certain properties of the dimensionality-reducing matrix M. Our work extends traditional results in sparse recovery, where instance optimality is expressed with respect to the set of sparse vectors, by replacing this set with an arbitrary finite union of subspaces. We show that the suggested instance optimality is equivalent to a generalized null space property of M and discuss possible relations with generalized restricted isometry properties.
{"title":"Compressed sensing and best approximation from unions of subspaces: Beyond dictionaries","authors":"Tomer Peleg, R. Gribonval, M. Davies","doi":"10.5281/ZENODO.43337","DOIUrl":"https://doi.org/10.5281/ZENODO.43337","url":null,"abstract":"We propose a theoretical study of the conditions guaranteeing that a decoder will obtain an optimal signal recovery from an underdetermined set of linear measurements. This special type of performance guarantee is termed instance optimality and is typically related with certain properties of the dimensionality-reducing matrix M. Our work extends traditional results in sparse recovery, where instance optimality is expressed with respect to the set of sparse vectors, by replacing this set with an arbitrary finite union of subspaces. We show that the suggested instance optimality is equivalent to a generalized null space property of M and discuss possible relations with generalized restricted isometry properties.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126451631","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper considers the watermark embedding problem onto Compressive Sensed measurements of a signal that is sparse in a proper basis. We propose a novel watermark encoding-decoding algorithm that exploits the sparsity of the signal to achieve dense watermarking. The proposed algorithm is robust under additive white Gaussian noise as well as impulsive noise or their mixture. The experimental results show also that the algorithm achieves an embedding capacity superior to those of classical ℓ2 and ℓ1 embedding algorithms.
{"title":"Robust watermarking of compressive sensed measurements under impulsive and Gaussian attacks","authors":"Mehmet Yamaç, Çagatay Dikici, B. Sankur","doi":"10.5281/ZENODO.43699","DOIUrl":"https://doi.org/10.5281/ZENODO.43699","url":null,"abstract":"This paper considers the watermark embedding problem onto Compressive Sensed measurements of a signal that is sparse in a proper basis. We propose a novel watermark encoding-decoding algorithm that exploits the sparsity of the signal to achieve dense watermarking. The proposed algorithm is robust under additive white Gaussian noise as well as impulsive noise or their mixture. The experimental results show also that the algorithm achieves an embedding capacity superior to those of classical ℓ2 and ℓ1 embedding algorithms.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134176373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sensor data fusion techniques have been applied in the recent years to the combination of the information provided by different sensor systems. Passive coherent location (PCL) networks use the illumination by common radio or television transmitters to detect air-targets and estimate their positions and parameters due to the reflected waves. To fuse the information of the bistatic Tx-Rx pairs advanced techniques have been developed based on the detections and parameter estimates obtained at each bistatic pair. In our paper we will consider joined signal processing of the radar raw data based on compressive sensing (CS) techniques using the block-sparsity approach.
{"title":"A compressive sensing approach to the fusion of PCL sensors","authors":"J. Ender","doi":"10.5281/ZENODO.43425","DOIUrl":"https://doi.org/10.5281/ZENODO.43425","url":null,"abstract":"Sensor data fusion techniques have been applied in the recent years to the combination of the information provided by different sensor systems. Passive coherent location (PCL) networks use the illumination by common radio or television transmitters to detect air-targets and estimate their positions and parameters due to the reflected waves. To fuse the information of the bistatic Tx-Rx pairs advanced techniques have been developed based on the detections and parameter estimates obtained at each bistatic pair. In our paper we will consider joined signal processing of the radar raw data based on compressive sensing (CS) techniques using the block-sparsity approach.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126597320","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper presents a fast local disparity calculation algorithm on calibrated stereo images for automotive applications. By utilizing the ground obstacle assumption for a typical road scene, only a small fraction of disparity space is required to be visited in order to find a disparity map. It works by using the neighbourhood disparities of the pixels in the lower image line as supporting points to determine the search range of its upper vicinity line. Unlike the conventional seed growing based algorithms that are only capable of producing a semi-dense disparity map, the proposed algorithm utilises information provided by each pixel rather than trusting only the featured seeds. Hence, it is capable of providing a denser disparity output with low errors in homogeneous areas. The experimental results are also compared to a normal exhaustive search (block matching) algorithm, showing a factor of ten improvement in speed, whilst the accuracy is enhanced by 20% without constraint to the maximum possible disparity.
{"title":"Efficient disparity calculation based on stereo vision with ground obstacle assumption","authors":"Zhen Zhang, X. Ai, N. Dahnoun","doi":"10.5281/ZENODO.43457","DOIUrl":"https://doi.org/10.5281/ZENODO.43457","url":null,"abstract":"This paper presents a fast local disparity calculation algorithm on calibrated stereo images for automotive applications. By utilizing the ground obstacle assumption for a typical road scene, only a small fraction of disparity space is required to be visited in order to find a disparity map. It works by using the neighbourhood disparities of the pixels in the lower image line as supporting points to determine the search range of its upper vicinity line. Unlike the conventional seed growing based algorithms that are only capable of producing a semi-dense disparity map, the proposed algorithm utilises information provided by each pixel rather than trusting only the featured seeds. Hence, it is capable of providing a denser disparity output with low errors in homogeneous areas. The experimental results are also compared to a normal exhaustive search (block matching) algorithm, showing a factor of ten improvement in speed, whilst the accuracy is enhanced by 20% without constraint to the maximum possible disparity.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124113740","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Speaker identification is a key component in many practical applications and the need of finding algorithms, which are robust under adverse noisy conditions, is extremely important. In this paper, the problem of text-independent speaker identification is studied in light of classification based on sparsity representation combined with a discriminative dictionary learning technique. Experimental evaluations on a small dataset reveal that the proposed method achieves a superior performance under short training sessions restrictions. In specific, the proposed method achieved high robustness for all the noisy conditions that were examined, when compared with a GMM universal background model (UBM-GMM) and sparse representation classification (SRC) approaches.
{"title":"Sparsity based robust speaker identification using a discriminative dictionary learning approach","authors":"Christos Tzagkarakis, A. Mouchtaris","doi":"10.5281/ZENODO.43348","DOIUrl":"https://doi.org/10.5281/ZENODO.43348","url":null,"abstract":"Speaker identification is a key component in many practical applications and the need of finding algorithms, which are robust under adverse noisy conditions, is extremely important. In this paper, the problem of text-independent speaker identification is studied in light of classification based on sparsity representation combined with a discriminative dictionary learning technique. Experimental evaluations on a small dataset reveal that the proposed method achieves a superior performance under short training sessions restrictions. In specific, the proposed method achieved high robustness for all the noisy conditions that were examined, when compared with a GMM universal background model (UBM-GMM) and sparse representation classification (SRC) approaches.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"117 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129575863","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The rational function systems proved to be useful in several areas including system and control theories and signal processing. In this paper, we present an extension of the well-known particle swarm optimization (PSO) method based on the hyperbolic geometry. We applied this method on digital signals to determine the optimal parameters of the rational function systems. Our goal is to minimize the error between the approximation and the original signal while the poles of the system remain stable. Namely, we show that the presented algorithm is suitable to localize the same poles by using different initial conditions.
{"title":"Hyperbolic particle swarm optimization with application in rational identification","authors":"P. Kovács, S. Kiranyaz, M. Gabbouj","doi":"10.5281/ZENODO.43626","DOIUrl":"https://doi.org/10.5281/ZENODO.43626","url":null,"abstract":"The rational function systems proved to be useful in several areas including system and control theories and signal processing. In this paper, we present an extension of the well-known particle swarm optimization (PSO) method based on the hyperbolic geometry. We applied this method on digital signals to determine the optimal parameters of the rational function systems. Our goal is to minimize the error between the approximation and the original signal while the poles of the system remain stable. Namely, we show that the presented algorithm is suitable to localize the same poles by using different initial conditions.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127497204","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper we propose a new technique for automatic correction of eye blink artifact in single channel EEG recording. The proposed technique consists of three steps. In the first two steps a dictionary matrix and a reference signal to the eye blink artifact are constructed from the recorded data, respectively. In the proposed technique we suggest building the dictionary matrix using empirical mode decomposition (EMD). In the last step, orthogonal matching pursuit (OMP) is utilized to find the minimum number of columns of the constructed dictionary matrix that fit the reference signal. Simulation results on real EEG data show that the proposed technique outperforms some of the existing single channel blind source separation techniques.
{"title":"Automatic correction of eye blink artifact in single channel EEG recording using EMD and OMP","authors":"N. Mourad, R. Niazy","doi":"10.5281/ZENODO.43591","DOIUrl":"https://doi.org/10.5281/ZENODO.43591","url":null,"abstract":"In this paper we propose a new technique for automatic correction of eye blink artifact in single channel EEG recording. The proposed technique consists of three steps. In the first two steps a dictionary matrix and a reference signal to the eye blink artifact are constructed from the recorded data, respectively. In the proposed technique we suggest building the dictionary matrix using empirical mode decomposition (EMD). In the last step, orthogonal matching pursuit (OMP) is utilized to find the minimum number of columns of the constructed dictionary matrix that fit the reference signal. Simulation results on real EEG data show that the proposed technique outperforms some of the existing single channel blind source separation techniques.","PeriodicalId":400766,"journal":{"name":"21st European Signal Processing Conference (EUSIPCO 2013)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122372097","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：