Pub Date : 2004-08-02DOI: 10.1109/NNSP.2003.1318040
Junying Zhang, Le Wei, Y. Wang
As a common feature in microarray profiling, gene expression profiles represent a composite of more than one distinct sources, which can severely decrease the sensitivity and specificity for the measurement of molecular signatures associated with different disease processes. Independent component analysis (ICA) is a widely applicable approach in blind source separation (BSS) but with limitations that the sources are independent, while a more common situation, which still happens in microarray profiles, is BSS where sources are not statistically independent. A novel idea of BSS is presented: it is a matrix factorization problem without enforcement of statistical characteristics on sources, while blind independent source separation is in fact matrix factorization, to factorize the observation matrix into a mixing matrix and a source matrix where the sources are independent. Since non-negative sources are meaningful in many applications including microarray profiling, we presented that blind non-negative source separation is essentially a matrix factorization, to factorize the observation matrix into a non-negative mixing matrix and a non-negative source matrix where the sources may be dependent. Non-negative matrix factorization (NMF) technique is applied to this non-negative source separation and is proven by a large number of computer simulations and by partial volume correction (PVC) experiments for real microarray data that it is effective when the sources are dependent with each other and/or are Gaussian distributed.
{"title":"Computational decomposition of molecular signatures based on blind source separation of non-negative dependent sources with NMF","authors":"Junying Zhang, Le Wei, Y. Wang","doi":"10.1109/NNSP.2003.1318040","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318040","url":null,"abstract":"As a common feature in microarray profiling, gene expression profiles represent a composite of more than one distinct sources, which can severely decrease the sensitivity and specificity for the measurement of molecular signatures associated with different disease processes. Independent component analysis (ICA) is a widely applicable approach in blind source separation (BSS) but with limitations that the sources are independent, while a more common situation, which still happens in microarray profiles, is BSS where sources are not statistically independent. A novel idea of BSS is presented: it is a matrix factorization problem without enforcement of statistical characteristics on sources, while blind independent source separation is in fact matrix factorization, to factorize the observation matrix into a mixing matrix and a source matrix where the sources are independent. Since non-negative sources are meaningful in many applications including microarray profiling, we presented that blind non-negative source separation is essentially a matrix factorization, to factorize the observation matrix into a non-negative mixing matrix and a non-negative source matrix where the sources may be dependent. Non-negative matrix factorization (NMF) technique is applied to this non-negative source separation and is proven by a large number of computer simulations and by partial volume correction (PVC) experiments for real microarray data that it is effective when the sources are dependent with each other and/or are Gaussian distributed.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2004-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127447990","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}
Pub Date : 2003-09-17DOI: 10.1109/NNSP.2003.1318030
Y. Rao, Deniz Erdoğmuş, G. Y. Rao, J. Príncipe
Linear system identification with noisy inputs is a critical problem in signal processing and control. Conventional techniques based on the mean squared-error (MSE) criterion can at best provide a biased estimate of the unknown system being modeled. Recently, we proposed a new criterion called the error whitening criterion (EWC) to solve the problem of linear parameter estimation in the presence of additive white noise. In this paper, we present a fixed-point type algorithm with O(N/sup 2/) complexity for EWC, called the recursive error whitening (REW) algorithm. We would also show that the EWC solution could be solved using the computational principles of total least squares (TLS). A novel EWC-TLS algorithm with O(N/sup 2/) complexity is derived. We will then apply the EWC methods for adaptive inverse control and show the superiority over existing methods.
{"title":"Fast error whitening algorithms for system identification and control","authors":"Y. Rao, Deniz Erdoğmuş, G. Y. Rao, J. Príncipe","doi":"10.1109/NNSP.2003.1318030","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318030","url":null,"abstract":"Linear system identification with noisy inputs is a critical problem in signal processing and control. Conventional techniques based on the mean squared-error (MSE) criterion can at best provide a biased estimate of the unknown system being modeled. Recently, we proposed a new criterion called the error whitening criterion (EWC) to solve the problem of linear parameter estimation in the presence of additive white noise. In this paper, we present a fixed-point type algorithm with O(N/sup 2/) complexity for EWC, called the recursive error whitening (REW) algorithm. We would also show that the EWC solution could be solved using the computational principles of total least squares (TLS). A novel EWC-TLS algorithm with O(N/sup 2/) complexity is derived. We will then apply the EWC methods for adaptive inverse control and show the superiority over existing methods.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"2016 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127476398","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}
Pub Date : 2003-09-17DOI: 10.1109/NNSP.2003.1318009
C. Jamet, S. Thiria, C. Moulin, M. Crépon
This paper presents a neuro-variational method to invert satellite ocean color signal. The method is based on a combination of neural networks and classical variational inversion. The radiative transfer equations are modeled by neural networks whose input are the oceanic and atmospheric parameters and output the top of the atmosphere reflectance at several wavelengths. The procedure consists in minimizing a quadratic cost function which is the distance between the satellite observed reflectance and the neural network computed reflectance, the control parameters being the oceanic and atmospheric parameters. The method allows us to retrieve atmospheric and oceanic parameters. We present a feasibility experiment. We show we can retrieve Chl-a with an error of 19.7% if we can obtain a perfect knowledge of three atmospheric parameters. Finally, an inversion of one SeaWiFS image is presented. The Chl-a give coherent spatial structures.
{"title":"Neuro-variational inversion of ocean color imagery","authors":"C. Jamet, S. Thiria, C. Moulin, M. Crépon","doi":"10.1109/NNSP.2003.1318009","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318009","url":null,"abstract":"This paper presents a neuro-variational method to invert satellite ocean color signal. The method is based on a combination of neural networks and classical variational inversion. The radiative transfer equations are modeled by neural networks whose input are the oceanic and atmospheric parameters and output the top of the atmosphere reflectance at several wavelengths. The procedure consists in minimizing a quadratic cost function which is the distance between the satellite observed reflectance and the neural network computed reflectance, the control parameters being the oceanic and atmospheric parameters. The method allows us to retrieve atmospheric and oceanic parameters. We present a feasibility experiment. We show we can retrieve Chl-a with an error of 19.7% if we can obtain a perfect knowledge of three atmospheric parameters. Finally, an inversion of one SeaWiFS image is presented. The Chl-a give coherent spatial structures.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114445085","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}
Pub Date : 2003-09-17DOI: 10.1109/NNSP.2003.1318021
A. Ypma, T. Heskes
We formulate extended Kalman smoothing in an expectation-propagation (EP) framework. The approximation involved (a local linearization) can be looked upon as a 'collapse' of a non-Gaussian belief state onto a Gaussian form. This formulation allows us to come up with better approximations to the belief states, since we can iterate the algorithm until no further refinement of the beliefs is obtained. Compared to the standard extended Kalman smoother, we linearize around the mode of the actual two-slice belief state instead of the predicted mean of the one-slice belief. In initial experiments with a one-dimensional nonlinear dynamical system we found that our method improves over the extended Kalman filter and performs comparable to the unscented Kalman filter, whereas only second-order approximations are being made. The EP-formulation in principle allows for incorporation of higher-order approximations, possibly leading to further improvements.
{"title":"Iterated extended Kalman smoothing with expectation-propagation","authors":"A. Ypma, T. Heskes","doi":"10.1109/NNSP.2003.1318021","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318021","url":null,"abstract":"We formulate extended Kalman smoothing in an expectation-propagation (EP) framework. The approximation involved (a local linearization) can be looked upon as a 'collapse' of a non-Gaussian belief state onto a Gaussian form. This formulation allows us to come up with better approximations to the belief states, since we can iterate the algorithm until no further refinement of the beliefs is obtained. Compared to the standard extended Kalman smoother, we linearize around the mode of the actual two-slice belief state instead of the predicted mean of the one-slice belief. In initial experiments with a one-dimensional nonlinear dynamical system we found that our method improves over the extended Kalman filter and performs comparable to the unscented Kalman filter, whereas only second-order approximations are being made. The EP-formulation in principle allows for incorporation of higher-order approximations, possibly leading to further improvements.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129285632","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}
Pub Date : 2003-09-17DOI: 10.1109/NNSP.2003.1318005
Neil D. Lawrence, M. Milo, M. Niranjan, P. Rashbass, S. Soullier
Gene expression measurements quantify the level of mRNA produced from each gene. Two principal methods exist for producing slides for extracting these levels: photolithography and spotted arrays. One difficulty with the spotted array format is determining the size and location of the spots on the array. In this paper we present a Bayesian approach to processing images produced by these arrays that seeks posterior distributions over the size and positions of the spots. This enables us to estimate expression ratios and their variances. Exact inference for the model we specify is intractable; we develop an approximate inference technique, which combines importance sampling with variational inference. Our technique has already been shown to be more consistent than both manual processing and another automated technique [N. D. Lawrence, et al., "Reducing the Variability in cDNA Microarray Image Processing by Inference"]. Here we present large-scale results for twenty-four microarray slides each representing 5760 genes and show the dramatic effects of incorporating variance in our downstream analysis. Software based on this algorithm is available for academic use.
{"title":"Bayesian processing of microarray images","authors":"Neil D. Lawrence, M. Milo, M. Niranjan, P. Rashbass, S. Soullier","doi":"10.1109/NNSP.2003.1318005","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318005","url":null,"abstract":"Gene expression measurements quantify the level of mRNA produced from each gene. Two principal methods exist for producing slides for extracting these levels: photolithography and spotted arrays. One difficulty with the spotted array format is determining the size and location of the spots on the array. In this paper we present a Bayesian approach to processing images produced by these arrays that seeks posterior distributions over the size and positions of the spots. This enables us to estimate expression ratios and their variances. Exact inference for the model we specify is intractable; we develop an approximate inference technique, which combines importance sampling with variational inference. Our technique has already been shown to be more consistent than both manual processing and another automated technique [N. D. Lawrence, et al., \"Reducing the Variability in cDNA Microarray Image Processing by Inference\"]. Here we present large-scale results for twenty-four microarray slides each representing 5760 genes and show the dramatic effects of incorporating variance in our downstream analysis. Software based on this algorithm is available for academic use.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"291 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124182227","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}
Pub Date : 2003-09-17DOI: 10.1109/NNSP.2003.1318043
G. Morison, T. Durrani
In this paper the problem of single input multiple output (SIMO) and multiple input multiple output (MIMO) blind equalization in a frequency selective environment is addressed using blind source separation techniques. A robust whitening stage is included to reduce the effects of noise enhancement that traditional prewhitening methods suffer, and the use of matrix momentum with the natural gradient algorithm is utilised to improve the computation efficiency of the standard natural gradient algorithm. The performance of the algorithm is demonstrated for an ill conditioned channel and compared with a current natural gradient based blind equalization using source separation method.
{"title":"Blind equalization using matrix momentum and natural gradient adaptation","authors":"G. Morison, T. Durrani","doi":"10.1109/NNSP.2003.1318043","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318043","url":null,"abstract":"In this paper the problem of single input multiple output (SIMO) and multiple input multiple output (MIMO) blind equalization in a frequency selective environment is addressed using blind source separation techniques. A robust whitening stage is included to reduce the effects of noise enhancement that traditional prewhitening methods suffer, and the use of matrix momentum with the natural gradient algorithm is utilised to improve the computation efficiency of the standard natural gradient algorithm. The performance of the algorithm is demonstrated for an ill conditioned channel and compared with a current natural gradient based blind equalization using source separation method.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121719612","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}
Pub Date : 2003-09-17DOI: 10.1109/NNSP.2003.1318001
Y. Wang, Junying Zhang, Javed I. Khan, R. Clarke, Zhiping Gu
Gene microarray technologies provide powerful tools for the large scale analysis of gene expression in cancer research. Clinical applications often aim to facilitate a molecular classification of cancers based on discriminatory genes associated with different clinical stages or outcomes. However, gene expression profiles often represent a composite of more than one distinct source due to tissue heterogeneity, and could result in extracting signatures reflecting the proportion of stromal contamination in the sample, rather than underlying tumor biology. We therefore wish to introduce a computational approach, which allows for a blind decomposition of gene expression profiles from mixed cell populations. The algorithm is based on a linear latent variable model, whose parameters are estimated using partially-independent component analysis, supported by a subset of differentially-expressed genes. We demonstrate the principle of the approach on the data sets derived from mixed cell lines of small round blue cell tumors. Because accurate source separation can be achieved blindly and numerically, we anticipate that computational correction of tissue heterogeneity would be useful in a wide variety of gene microarray studies.
{"title":"Partially-independent component analysis for tissue heterogeneity correction in microarray gene expression analysis","authors":"Y. Wang, Junying Zhang, Javed I. Khan, R. Clarke, Zhiping Gu","doi":"10.1109/NNSP.2003.1318001","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318001","url":null,"abstract":"Gene microarray technologies provide powerful tools for the large scale analysis of gene expression in cancer research. Clinical applications often aim to facilitate a molecular classification of cancers based on discriminatory genes associated with different clinical stages or outcomes. However, gene expression profiles often represent a composite of more than one distinct source due to tissue heterogeneity, and could result in extracting signatures reflecting the proportion of stromal contamination in the sample, rather than underlying tumor biology. We therefore wish to introduce a computational approach, which allows for a blind decomposition of gene expression profiles from mixed cell populations. The algorithm is based on a linear latent variable model, whose parameters are estimated using partially-independent component analysis, supported by a subset of differentially-expressed genes. We demonstrate the principle of the approach on the data sets derived from mixed cell lines of small round blue cell tumors. Because accurate source separation can be achieved blindly and numerically, we anticipate that computational correction of tissue heterogeneity would be useful in a wide variety of gene microarray studies.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130229688","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}
Pub Date : 2003-09-17DOI: 10.1109/NNSP.2003.1318013
P. Garda, O. Romain, B. Granado, A. Pinna, D. Faura, K. Hachicha
In this paper, we introduce the architecture of an intelligent beacon for wireless sensor networks. This beacon acquires images of a scene and detects motion, thanks to the real-time execution of a Markov motion detection algorithm. When some motion is detected, neural networks are applied in real-time to the acquired images in order to trigger some alarm. Finally, when some alarm is triggered, video of the scene compressed with the MMJPEG2000 algorithm are sent on a wireless network, a long-range communication by satellite for example. The beacon is implemented on a platform including a microcontroller, a DSP, an FPGA and several dedicated modules.
{"title":"Architecture of an intelligent beacon for wireless sensor networks","authors":"P. Garda, O. Romain, B. Granado, A. Pinna, D. Faura, K. Hachicha","doi":"10.1109/NNSP.2003.1318013","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318013","url":null,"abstract":"In this paper, we introduce the architecture of an intelligent beacon for wireless sensor networks. This beacon acquires images of a scene and detects motion, thanks to the real-time execution of a Markov motion detection algorithm. When some motion is detected, neural networks are applied in real-time to the acquired images in order to trigger some alarm. Finally, when some alarm is triggered, video of the scene compressed with the MMJPEG2000 algorithm are sent on a wireless network, a long-range communication by satellite for example. The beacon is implemented on a platform including a microcontroller, a DSP, an FPGA and several dedicated modules.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"82 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128596869","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}
Pub Date : 2003-09-17DOI: 10.1109/NNSP.2003.1318039
Dorothee E. Marossero, Deniz Erdoğmuş, N. Euliano, J. Príncipe, K. Hild
Fetal heart rate (FHR) monitoring is currently the primary methodology for antenatal determination of fetal well-being. Currently, the FHR can be detected with ultrasonography, but the additional information from fetal electrocardiogram (FECG) is only available via an invasive scalp electrode. A cost effective noninvasive monitoring through standard ECG electrodes could be used on nearly every patient in lieu of the ultrasound monitors. In this method, a number of electrodes are positioned on the abdomen of the mother to collect, simultaneously, various combinations of the signals including the heartbeats of the mother and the fetus. For accurate fetal heart-rate estimation, a clean FECG must be extracted from the collected mixtures. It is well known that this can be achieved using blind source separation (BSS) techniques. In this paper, the performance of the Mermaid algorithm, which is based on minimizing Renyi's mutual information, is evaluated on this problem of great practical importance. The effectiveness and data efficiency of Mermaid and its superiority over alternative information theoretic BSS algorithms are illustrated using artificially mixed ECG signals as well as fetal heart rate estimates in real ECG mixtures.
{"title":"Independent components analysis for fetal electrocardiogram extraction: a case for the data efficient Mermaid algorithm","authors":"Dorothee E. Marossero, Deniz Erdoğmuş, N. Euliano, J. Príncipe, K. Hild","doi":"10.1109/NNSP.2003.1318039","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318039","url":null,"abstract":"Fetal heart rate (FHR) monitoring is currently the primary methodology for antenatal determination of fetal well-being. Currently, the FHR can be detected with ultrasonography, but the additional information from fetal electrocardiogram (FECG) is only available via an invasive scalp electrode. A cost effective noninvasive monitoring through standard ECG electrodes could be used on nearly every patient in lieu of the ultrasound monitors. In this method, a number of electrodes are positioned on the abdomen of the mother to collect, simultaneously, various combinations of the signals including the heartbeats of the mother and the fetus. For accurate fetal heart-rate estimation, a clean FECG must be extracted from the collected mixtures. It is well known that this can be achieved using blind source separation (BSS) techniques. In this paper, the performance of the Mermaid algorithm, which is based on minimizing Renyi's mutual information, is evaluated on this problem of great practical importance. The effectiveness and data efficiency of Mermaid and its superiority over alternative information theoretic BSS algorithms are illustrated using artificially mixed ECG signals as well as fetal heart rate estimates in real ECG mixtures.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125610215","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}
Pub Date : 2003-09-17DOI: 10.1109/NNSP.2003.1318011
J. Delvit, D. Léger, S. Roques, C. Valorge
In the context of Earth observation satellites such as SPOT or IKONOS, it is important to measure the modulation transfer function (MTF) and the noise in order to quantify the quality of the imaging system. This measurement is useful to decide to focus the telescope or to make a deconvolution filter whose purpose is to enhance image contrast. This paper presents a univariant MTF and noise measurement method using non specific views. It is a particular application of a general approach of image quality assessment. The method presented in this paper is based on artificial neural network (ANN) use. The ANN learns how to recognize MTF and noise from known images, and the neural network is able, after the learning step, to assess the MTF and the noise from unknown images.
{"title":"Modulation transfer function and noise measurement using neural networks","authors":"J. Delvit, D. Léger, S. Roques, C. Valorge","doi":"10.1109/NNSP.2003.1318011","DOIUrl":"https://doi.org/10.1109/NNSP.2003.1318011","url":null,"abstract":"In the context of Earth observation satellites such as SPOT or IKONOS, it is important to measure the modulation transfer function (MTF) and the noise in order to quantify the quality of the imaging system. This measurement is useful to decide to focus the telescope or to make a deconvolution filter whose purpose is to enhance image contrast. This paper presents a univariant MTF and noise measurement method using non specific views. It is a particular application of a general approach of image quality assessment. The method presented in this paper is based on artificial neural network (ANN) use. The ANN learns how to recognize MTF and noise from known images, and the neural network is able, after the learning step, to assess the MTF and the noise from unknown images.","PeriodicalId":315958,"journal":{"name":"2003 IEEE XIII Workshop on Neural Networks for Signal Processing (IEEE Cat. No.03TH8718)","volume":"37 19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125704883","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：