For the convolutive mixture, a subspace method to separate the sources is proposed. It is showed that after using only the second order statistic but more sensors than sources, the convolutive mixture can be itentified up to instantaneou mixture. Furthermore, the sources can be separated by any algorithm for instantaneous mixture (based in generally on the fourth order statistics).
{"title":"Subspace method for blind separation of sources in convolutive mixture","authors":"A. Mansour, C. Jutten, P. Loubaton","doi":"10.5281/ZENODO.35953","DOIUrl":"https://doi.org/10.5281/ZENODO.35953","url":null,"abstract":"For the convolutive mixture, a subspace method to separate the sources is proposed. It is showed that after using only the second order statistic but more sensors than sources, the convolutive mixture can be itentified up to instantaneou mixture. Furthermore, the sources can be separated by any algorithm for instantaneous mixture (based in generally on the fourth order statistics).","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"343 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124314383","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, a computational metric that incorporates many aspects of human vision and color perception to predict the quality of color coded images is presented. The proposed distortion measure is built on opponent-colors theory and on a multi-channel model of spatial vision. The metric has been validated by psychophysical data on 400 images and two human observers.
{"title":"Perceptual quality metric for digitally coded color images","authors":"C. Lambrecht, J. Farrell","doi":"10.5281/ZENODO.36035","DOIUrl":"https://doi.org/10.5281/ZENODO.36035","url":null,"abstract":"In this paper, a computational metric that incorporates many aspects of human vision and color perception to predict the quality of color coded images is presented. The proposed distortion measure is built on opponent-colors theory and on a multi-channel model of spatial vision. The metric has been validated by psychophysical data on 400 images and two human observers.","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"105 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124048785","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 monocular vision, a priori knowledge is necessary to perform 3D reconstruction. This paper describes how to evaluate two out of six external parameters of a camera in order to project an image on a curved surface (generalized cylinder). The final aim consists of reconstructing the model of the surface. Afterwards, with this model we can derive a flat representation of the scene without any distortions due to the projective geometry. In this work based on one projected view of the scene, we develop two methods to detect the projection of the revolution axis of the curved surface. With this axis, we can then extract the external parameters of a camera. The first one is based on the derivation of a polynomial function and the second one is based on the detection of the common normal between curves.
{"title":"Curved surface reconstruction using monocular vision","authors":"W. Puech, J. Chassery","doi":"10.5281/ZENODO.36323","DOIUrl":"https://doi.org/10.5281/ZENODO.36323","url":null,"abstract":"In monocular vision, a priori knowledge is necessary to perform 3D reconstruction. This paper describes how to evaluate two out of six external parameters of a camera in order to project an image on a curved surface (generalized cylinder). The final aim consists of reconstructing the model of the surface. Afterwards, with this model we can derive a flat representation of the scene without any distortions due to the projective geometry. In this work based on one projected view of the scene, we develop two methods to detect the projection of the revolution axis of the curved surface. With this axis, we can then extract the external parameters of a camera. The first one is based on the derivation of a polynomial function and the second one is based on the detection of the common normal between curves.","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126385004","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}
Voiceband channels are frequently used for data transmission, even though they were not designed for such a use. The reason is very simple; such channels already exist. It is also clear that such channels when used for data transmission can not be used at the same time for voice transmission, and vice versa. However, there are a lot of applications where simultaneous transmission of voice and data through the existing voiceband channel is needed. In this work we propose a method for simultaneous transmission based on orthogonalization of the voice signal. A comprehensive assessment of the orthogonal voice which includes subjective measures shows that the orthogonal signal may have full intelligibility while its quality is only slightly degraded. The MOS for orthogonal voice is in the range 2.5–3.9 and depend on the data transmission parameters.
{"title":"Using orthogonalized voice for simultaneous transmission of voice and data","authors":"M. Goren, O. Tirosh, L. Kishon-Rabin, D. Wulich","doi":"10.5281/ZENODO.36176","DOIUrl":"https://doi.org/10.5281/ZENODO.36176","url":null,"abstract":"Voiceband channels are frequently used for data transmission, even though they were not designed for such a use. The reason is very simple; such channels already exist. It is also clear that such channels when used for data transmission can not be used at the same time for voice transmission, and vice versa. However, there are a lot of applications where simultaneous transmission of voice and data through the existing voiceband channel is needed. In this work we propose a method for simultaneous transmission based on orthogonalization of the voice signal. A comprehensive assessment of the orthogonal voice which includes subjective measures shows that the orthogonal signal may have full intelligibility while its quality is only slightly degraded. The MOS for orthogonal voice is in the range 2.5–3.9 and depend on the data transmission parameters.","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"40 12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128169036","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}
T. Isernia, V. Pascazio, R. Pierri, Gilda Schirinzi
A new method to accurately reconstruct a Synthetic Aperture Radar complex image starting from phase errors affected raw received data is presented. It is based on a phase retrieval algorithm, and the unknown complex reflectivity is found by minimising a proper functional using die partial phase information carried out by the phase corrupted raw data as the initial guess of an iterative procedure. The method, which is capable of compensating for both 1-D and 2-D phase errors, has been validated on real
{"title":"SAR images reconstruction via phase retrieval","authors":"T. Isernia, V. Pascazio, R. Pierri, Gilda Schirinzi","doi":"10.5281/ZENODO.36311","DOIUrl":"https://doi.org/10.5281/ZENODO.36311","url":null,"abstract":"A new method to accurately reconstruct a Synthetic Aperture Radar complex image starting from phase errors affected raw received data is presented. It is based on a phase retrieval algorithm, and the unknown complex reflectivity is found by minimising a proper functional using die partial phase information carried out by the phase corrupted raw data as the initial guess of an iterative procedure. The method, which is capable of compensating for both 1-D and 2-D phase errors, has been validated on real <laia.","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131337289","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 work presents training methods and recognition experiments for phoneme-wise tied mixture densities in hidden Markov models (HMM). The system trains speaker dependent, but vocabulary independent, phoneme models for the recognition of Finnish words. The Learning Vector Quantization (LVQ) methods are applied to increase the discrimination between the phoneme models. A segmental LVQ3 training is proposed to substitute the LVQ2 based corrective tuning as a parameter estimation method. The experiments indicate that the new method can provide the corresponding recognition accuracy, but with less training and more robustness over the initial models. Experiments to upscale the current system by introducing context vectors and larger mixture pools show up to 40 % reduction of recognition errors compared to the earlier results in [10].
{"title":"Segmental LVQ3 training for phoneme-wise tied mixture density HMMS","authors":"M. Kurimo","doi":"10.5281/ZENODO.36307","DOIUrl":"https://doi.org/10.5281/ZENODO.36307","url":null,"abstract":"This work presents training methods and recognition experiments for phoneme-wise tied mixture densities in hidden Markov models (HMM). The system trains speaker dependent, but vocabulary independent, phoneme models for the recognition of Finnish words. The Learning Vector Quantization (LVQ) methods are applied to increase the discrimination between the phoneme models. A segmental LVQ3 training is proposed to substitute the LVQ2 based corrective tuning as a parameter estimation method. The experiments indicate that the new method can provide the corresponding recognition accuracy, but with less training and more robustness over the initial models. Experiments to upscale the current system by introducing context vectors and larger mixture pools show up to 40 % reduction of recognition errors compared to the earlier results in [10].","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"55 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132117862","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}
Linear constraints in adaptive beamformer are often used to control its transfer function. In this paper we utilized these constraints lo reduce computational cost of the adaptive algorithm. For this aim. two types of constraints were proposed, The first one is that all zeros of the transfer function appear as conjugate-complex pairs lying on the unit circle. The second one is that some zeros have prescribed positions and the adaptation is realized by the rest of zeros. Developed constraints are applied to the generalized sidelobe canceller and used to blocking matrix design. Experiments proved that degradation in performance of the partially adaptive algorithm is a little compared to the Tuli adaptive algorithm.
{"title":"Partially adaptive generalized sidelobe canceller with prescribed zeros","authors":"Z. Šarić, Milorad Cetina","doi":"10.5281/zenodo.36136","DOIUrl":"https://doi.org/10.5281/zenodo.36136","url":null,"abstract":"Linear constraints in adaptive beamformer are often used to control its transfer function. In this paper we utilized these constraints lo reduce computational cost of the adaptive algorithm. For this aim. two types of constraints were proposed, The first one is that all zeros of the transfer function appear as conjugate-complex pairs lying on the unit circle. The second one is that some zeros have prescribed positions and the adaptation is realized by the rest of zeros. Developed constraints are applied to the generalized sidelobe canceller and used to blocking matrix design. Experiments proved that degradation in performance of the partially adaptive algorithm is a little compared to the Tuli adaptive algorithm.","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132561414","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 addresses the classification of multispectral remote-sensing images by the neural-network approach. In particular, an experimental comparison on the performances provided by different neural models for classifying multisensor remote-sensing data is reported. Four neural classifiers are considered in the comparison: the Multilayer Perceptron, Probabilistic Neural Networks, Radial Basis Function networks and a kind of Structured Neural Networks.
{"title":"Classification of multispectral remote-sensing images by neural networks","authors":"F. Roli, S. Serpico, L. Bruzzone, G. Vernazza","doi":"10.5281/ZENODO.36388","DOIUrl":"https://doi.org/10.5281/ZENODO.36388","url":null,"abstract":"This paper addresses the classification of multispectral remote-sensing images by the neural-network approach. In particular, an experimental comparison on the performances provided by different neural models for classifying multisensor remote-sensing data is reported. Four neural classifiers are considered in the comparison: the Multilayer Perceptron, Probabilistic Neural Networks, Radial Basis Function networks and a kind of Structured Neural Networks.","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130090980","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, an approach to improve the spatial resolution of color video is presented. Such high resolution images are desired, for example, in video printing. Previous work has shown that the most important step in achieving high quality results is the accuracy of the motion field. It is well known that motion estimation is an ill-posed problem. However, in processing color video, additional information contained in the color channels may be used to improve the accuracy of the motion field over the motion field obtained with the use of only one channel. In turn, this improvement in the motion field will be shown through several experimental results to significantly improve the estimation of a high resolution image sequence from a corresponding observed low resolution sequence.
{"title":"Resolution enhancement of color video","authors":"B. C. Tom, A. Katsaggelos","doi":"10.5281/ZENODO.36381","DOIUrl":"https://doi.org/10.5281/ZENODO.36381","url":null,"abstract":"In this paper, an approach to improve the spatial resolution of color video is presented. Such high resolution images are desired, for example, in video printing. Previous work has shown that the most important step in achieving high quality results is the accuracy of the motion field. It is well known that motion estimation is an ill-posed problem. However, in processing color video, additional information contained in the color channels may be used to improve the accuracy of the motion field over the motion field obtained with the use of only one channel. In turn, this improvement in the motion field will be shown through several experimental results to significantly improve the estimation of a high resolution image sequence from a corresponding observed low resolution sequence.","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129401922","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 derive expressions for the different linear and decision feedback equalizers in burst mode in the multichannel case. Among them we derive the class of unbiased minimum mean squared error equalizers. Optimal burst mode filters are found to be time-varying. Performance comparisons between these equalizers are done in terms of SNR and probability of error: these measures depend on the position in the burst. We study furthermore the performance when symbols are known or not at the edges of the burst and compare it to the continuous processing level. Finally we show that (time-invariant) continuous processing applied to burst mode can be organized to give sufficiently good performance, so that optimal (time-varying) burst processing implementation can be avoided.
{"title":"Unbiased MMSE decision-feedback equalization for packet transmission","authors":"D. Slock, E. Carvalho","doi":"10.5281/ZENODO.36187","DOIUrl":"https://doi.org/10.5281/ZENODO.36187","url":null,"abstract":"We derive expressions for the different linear and decision feedback equalizers in burst mode in the multichannel case. Among them we derive the class of unbiased minimum mean squared error equalizers. Optimal burst mode filters are found to be time-varying. Performance comparisons between these equalizers are done in terms of SNR and probability of error: these measures depend on the position in the burst. We study furthermore the performance when symbols are known or not at the edges of the burst and compare it to the continuous processing level. Finally we show that (time-invariant) continuous processing applied to burst mode can be organized to give sufficiently good performance, so that optimal (time-varying) burst processing implementation can be avoided.","PeriodicalId":282153,"journal":{"name":"1996 8th European Signal Processing Conference (EUSIPCO 1996)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133334730","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}