Pub Date : 2018-07-01DOI: 10.1109/iwobi.2018.8464189
{"title":"2018 IEEE International Work Conference on Bioinspired Intelligence","authors":"","doi":"10.1109/iwobi.2018.8464189","DOIUrl":"https://doi.org/10.1109/iwobi.2018.8464189","url":null,"abstract":"","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114564441","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 : 2018-07-01DOI: 10.1109/IWOBI.2018.8464204
Marvin Coto-Jiménez
Several attempts to enhance statistical parametric speech synthesis have contemplated deep-learning-based postfil-ters, which learn to perform a mapping of the synthetic speech parameters to the natural ones, reducing the gap between them. In this paper, we introduce a new pre-training approach for neural networks, applied in LSTM-based postfilters for speech synthesis, with the objective of enhancing the quality of the synthesized speech in a more efficient manner. Our approach begins with an auto-regressive training of one LSTM network, whose is used as an initialization for postfilters based on a denoising autoencoder architecture. We show the advantages of this initialization on a set of multi-stream postfilters, which encompass a collection of denoising autoencoders for the set of MFCC and fundamental frequency parameters of the artificial voice. Results show that the initialization succeeds in lowering the training time of the LSTM networks and achieves better results in enhancing the statistical parametric speech in most cases, when compared to the common random-initialized approach of the networks.
{"title":"Pre-training Long Short-term Memory Neural Networks for Efficient Regression in Artificial Speech Postfiltering","authors":"Marvin Coto-Jiménez","doi":"10.1109/IWOBI.2018.8464204","DOIUrl":"https://doi.org/10.1109/IWOBI.2018.8464204","url":null,"abstract":"Several attempts to enhance statistical parametric speech synthesis have contemplated deep-learning-based postfil-ters, which learn to perform a mapping of the synthetic speech parameters to the natural ones, reducing the gap between them. In this paper, we introduce a new pre-training approach for neural networks, applied in LSTM-based postfilters for speech synthesis, with the objective of enhancing the quality of the synthesized speech in a more efficient manner. Our approach begins with an auto-regressive training of one LSTM network, whose is used as an initialization for postfilters based on a denoising autoencoder architecture. We show the advantages of this initialization on a set of multi-stream postfilters, which encompass a collection of denoising autoencoders for the set of MFCC and fundamental frequency parameters of the artificial voice. Results show that the initialization succeeds in lowering the training time of the LSTM networks and achieves better results in enhancing the statistical parametric speech in most cases, when compared to the common random-initialized approach of the networks.","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123491887","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 : 2018-07-01DOI: 10.1109/IWOBI.2018.8464208
A. Mora-Zuniga, Steve Quiros-Barrantes, Francisco Siles
In this paper a workflow to extract cell features from brightfield microscopy image sequences is proposed. An event driven approach, combined with a forward and backward tracking limited by the cell's circularity was proven enough to extract relevant features that can be used to classify the cells into four phenotypes related to chemosensitivity studies: cell cycle arrest, apoptotic, damage proliferation and cells that have repaired their DNA damage. An average F1-Score greater than 0.7 was achieved in the detection and follow up of the events on images that present characteristics that impede the use of classic image segmentation and methods.
{"title":"M-Phase Feature Extraction Algorithm for Phenotype Classification from Cancer Brightfield Microscopy","authors":"A. Mora-Zuniga, Steve Quiros-Barrantes, Francisco Siles","doi":"10.1109/IWOBI.2018.8464208","DOIUrl":"https://doi.org/10.1109/IWOBI.2018.8464208","url":null,"abstract":"In this paper a workflow to extract cell features from brightfield microscopy image sequences is proposed. An event driven approach, combined with a forward and backward tracking limited by the cell's circularity was proven enough to extract relevant features that can be used to classify the cells into four phenotypes related to chemosensitivity studies: cell cycle arrest, apoptotic, damage proliferation and cells that have repaired their DNA damage. An average F1-Score greater than 0.7 was achieved in the detection and follow up of the events on images that present characteristics that impede the use of classic image segmentation and methods.","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125553419","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 : 2018-07-01DOI: 10.1109/IWOBI.2018.8464186
L. Chavarría-Zamora, Sergio Arriola-Valverde, R. Rímolo-Donadío
Haze is a natural scaterring effect of light that can blur images, lowering their quality and difficulting the scene analysis. In this paper, several state-of-the-art dehazing algorithms are evaluated through a subjective method for image quality assesment. The Dark Channel Prior algorithm, which resulted the best option within the evaluated algorithms, was further optimized by manipulating color spaces before its application in order to achieve a better reconstruction of the images.
{"title":"Evaluation of Fog Reduction Algorithms for Photogrammetric Applications in Agriculture","authors":"L. Chavarría-Zamora, Sergio Arriola-Valverde, R. Rímolo-Donadío","doi":"10.1109/IWOBI.2018.8464186","DOIUrl":"https://doi.org/10.1109/IWOBI.2018.8464186","url":null,"abstract":"Haze is a natural scaterring effect of light that can blur images, lowering their quality and difficulting the scene analysis. In this paper, several state-of-the-art dehazing algorithms are evaluated through a subjective method for image quality assesment. The Dark Channel Prior algorithm, which resulted the best option within the evaluated algorithms, was further optimized by manipulating color spaces before its application in order to achieve a better reconstruction of the images.","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131313058","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 : 2018-07-01DOI: 10.1109/IWOBI.2018.8464131
Anjali Yadav, M. Dutta, C. Travieso-González, J. B. Alonso
Cardiovascular diseases are very common these days and there arises a need for regular diagnosis of humans. Phonocardiogram is an effective diagnostic tool for analysing the heart sound. It helps in providing better information regarding clinical condition of the heart. This paper proposes an algorithmic method for differentiating a normal heart sound from an abnormal one using the PCG sound data. Cepstrum analysis has been performed on both types of signals and features are extracted from the heart sound. The extracted features are trained and tested with the help of a support vector machine classifier. The proposed method has achieved an accuracy of 95% in correctly classifying a heart sound PCG signal as normal and abnormal.
{"title":"Automatic Classification of Normal and Abnormal PCG Recording Heart Sound Recording Using Fourier Transform","authors":"Anjali Yadav, M. Dutta, C. Travieso-González, J. B. Alonso","doi":"10.1109/IWOBI.2018.8464131","DOIUrl":"https://doi.org/10.1109/IWOBI.2018.8464131","url":null,"abstract":"Cardiovascular diseases are very common these days and there arises a need for regular diagnosis of humans. Phonocardiogram is an effective diagnostic tool for analysing the heart sound. It helps in providing better information regarding clinical condition of the heart. This paper proposes an algorithmic method for differentiating a normal heart sound from an abnormal one using the PCG sound data. Cepstrum analysis has been performed on both types of signals and features are extracted from the heart sound. The extracted features are trained and tested with the help of a support vector machine classifier. The proposed method has achieved an accuracy of 95% in correctly classifying a heart sound PCG signal as normal and abnormal.","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123770838","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 : 2018-07-01DOI: 10.1109/IWOBI.2018.8464182
J. Vargas, R. Mora-Rodríguez, Francisco Siles
Cancer disease causes millions of deaths throughout the world and thousands in Costa Rica, and cancer treatment causes an immense economic burden on the social security system because drugs against the disease are extremely expensive. Through DNA sequencing techniques, the copy number of each gene could be found in order to describe how many times a gene is repeated within chromosomes. This type of data is of great importance since the cancer manifests a phenomenon called aneuploidy that consists of alterations in the number of chromosomes copies. Theories about the importance of aneuploidy in cancer supposed this phenomenon is a evolutionary engine that allows the disease to grow and resist changes produced by the organism and the treatments applied to these tissues. Studies have collected data on the number of copies in well-known databases such as CCLE and TCGA, which can be used to analyze the relationship between alterations in DNA and resistance to chemotherapies. In this study as a contribution, it was proposed to use statistical pattern recognition methods in order to identify chromosomal regions of DNA related to cancer chemosensitivity subtypes (resistant and sensitive subgroups) and then use these regions for CCLE cell lines labeling and TCGA samples classification.
{"title":"Genome Copy Number Feature Selection Based on Chromosomal Regions Alterations and Chemosensitivity Subtypes","authors":"J. Vargas, R. Mora-Rodríguez, Francisco Siles","doi":"10.1109/IWOBI.2018.8464182","DOIUrl":"https://doi.org/10.1109/IWOBI.2018.8464182","url":null,"abstract":"Cancer disease causes millions of deaths throughout the world and thousands in Costa Rica, and cancer treatment causes an immense economic burden on the social security system because drugs against the disease are extremely expensive. Through DNA sequencing techniques, the copy number of each gene could be found in order to describe how many times a gene is repeated within chromosomes. This type of data is of great importance since the cancer manifests a phenomenon called aneuploidy that consists of alterations in the number of chromosomes copies. Theories about the importance of aneuploidy in cancer supposed this phenomenon is a evolutionary engine that allows the disease to grow and resist changes produced by the organism and the treatments applied to these tissues. Studies have collected data on the number of copies in well-known databases such as CCLE and TCGA, which can be used to analyze the relationship between alterations in DNA and resistance to chemotherapies. In this study as a contribution, it was proposed to use statistical pattern recognition methods in order to identify chromosomal regions of DNA related to cancer chemosensitivity subtypes (resistant and sensitive subgroups) and then use these regions for CCLE cell lines labeling and TCGA samples classification.","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116463746","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 : 2018-07-01DOI: 10.1109/IWOBI.2018.8464217
Juan S. Ríos-Ramos, O. A. Nava, Hilda Maria Chable Martinez, Eduardo Rodríguez-Martínez
The high computational cost in computing disparity maps has relegated their use in real-time computer vision tasks. This work shows the design of a parallel algorithm for the generation of disparity maps which decreases the redundant operations during the stereo matching process. The proposed algorithm also uses a median filter to decrease the noise present in the generated disparity map, achieving an acceleration of 1135x on a GPU. In addition to the implementation of the parallel algorithm, GPU resources were used to improve memory access time and code optimization.
{"title":"Parallel Implementation in a GPU of the Calculation of Disparity Maps for Computer Vision","authors":"Juan S. Ríos-Ramos, O. A. Nava, Hilda Maria Chable Martinez, Eduardo Rodríguez-Martínez","doi":"10.1109/IWOBI.2018.8464217","DOIUrl":"https://doi.org/10.1109/IWOBI.2018.8464217","url":null,"abstract":"The high computational cost in computing disparity maps has relegated their use in real-time computer vision tasks. This work shows the design of a parallel algorithm for the generation of disparity maps which decreases the redundant operations during the stereo matching process. The proposed algorithm also uses a median filter to decrease the noise present in the generated disparity map, achieving an acceleration of 1135x on a GPU. In addition to the implementation of the parallel algorithm, GPU resources were used to improve memory access time and code optimization.","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131223821","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 : 2018-07-01DOI: 10.1109/IWOBI.2018.8464185
J. D. Zamora-Bolanos, M. Vílchez-Monge, Gabriela Ortiz-León, J. L. Crespo-Mariño
In Costa Rica, cardiovascular diseases are one of the principal death causes. At Instituto Tecnologico de Costa Rica there is a working group developing an axial ventricular assistance device. One of the problems that must be solved is the design of a system allowing the rotor to levitate within the device's channel, avoiding the impeller to cause any damage to the blood. This paper presents the design process for a passive magnetic suspension system prototype, based on the use of an external concentrator bearing, and an internal radial bearing.
{"title":"Preliminary Design Methodology and Prototype of a Passive Magnetic Suspension System for a Blood Axial Flow Pump","authors":"J. D. Zamora-Bolanos, M. Vílchez-Monge, Gabriela Ortiz-León, J. L. Crespo-Mariño","doi":"10.1109/IWOBI.2018.8464185","DOIUrl":"https://doi.org/10.1109/IWOBI.2018.8464185","url":null,"abstract":"In Costa Rica, cardiovascular diseases are one of the principal death causes. At Instituto Tecnologico de Costa Rica there is a working group developing an axial ventricular assistance device. One of the problems that must be solved is the design of a system allowing the rotor to levitate within the device's channel, avoiding the impeller to cause any damage to the blood. This paper presents the design process for a passive magnetic suspension system prototype, based on the use of an external concentrator bearing, and an internal radial bearing.","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127955805","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 : 2018-07-01DOI: 10.1109/IWOBI.2018.8464133
Peter Rot, Ž. Emeršič, V. Štruc, P. Peer
Segmentation techniques for ocular biometrics typically focus on finding a single eye region in the input image at the time. Only limited work has been done on multi-class eye segmentation despite a number of obvious advantages. In this paper we address this gap and present a deep multi-class eye segmentation model build around the SegNet architecture. We train the model on a small dataset (of 120 samples) of eye images and observe it to generalize well to unseen images and to ensure highly accurate segmentation results. We evaluate the model on the Multi-Angle Sclera Database (MASD) dataset and describe comprehensive experiments focusing on: i) segmentation performance, ii) error analysis, iii) the sensitivity of the model to changes in view direction, and iv) comparisons with competing single-class techniques. Our results show that the proposed model is viable solution for multi-class eye segmentation suitable for recognition (multi-biometric) pipelines based on ocular characteristics.
{"title":"Deep Multi-class Eye Segmentation for Ocular Biometrics","authors":"Peter Rot, Ž. Emeršič, V. Štruc, P. Peer","doi":"10.1109/IWOBI.2018.8464133","DOIUrl":"https://doi.org/10.1109/IWOBI.2018.8464133","url":null,"abstract":"Segmentation techniques for ocular biometrics typically focus on finding a single eye region in the input image at the time. Only limited work has been done on multi-class eye segmentation despite a number of obvious advantages. In this paper we address this gap and present a deep multi-class eye segmentation model build around the SegNet architecture. We train the model on a small dataset (of 120 samples) of eye images and observe it to generalize well to unseen images and to ensure highly accurate segmentation results. We evaluate the model on the Multi-Angle Sclera Database (MASD) dataset and describe comprehensive experiments focusing on: i) segmentation performance, ii) error analysis, iii) the sensitivity of the model to changes in view direction, and iv) comparisons with competing single-class techniques. Our results show that the proposed model is viable solution for multi-class eye segmentation suitable for recognition (multi-biometric) pipelines based on ocular characteristics.","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126850887","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 : 2018-07-01DOI: 10.1109/IWOBI.2018.8464219
A. Alvarado, Ricardo Corrales, Maria José Soares Leal, A. Ossa, R. Mora, Manuel Arroyo, Andrea Gomez, Alan Calderon, Jorge L. Arias-Arias
In this paper we present a computational model aimed at characterizing the Zika viral infection at a cellular level based on measurements done on viral Dulbecco plaques over time, and describe our current state of progress in the modeling task. So far we have developed an agent-based simulation model of the dispersion of the virus on the cells conforming the viral plaque. The growth rate of the viral plaques and the number of cells counted on each plaque were used to characterize the viral infection in terms of parameters related to the fate of infected cells, such as the probability of a cell infecting its neighboring cells and the probability of an infected cell of dying at any given moment. The model can be used to predict viral plaque growth patterns similar to those observed in the laboratory. Our current efforts focus on optimizing the model parameters to fit the experimental data. Further development of the model includes the description of viral infection kinetics of specific viral strains. Our model has been developed using the agent-based modeling language Netlogo [1].
{"title":"Cellular-Level Characterization of Dengue and Zika Virus Infection Using Multiagent Simulation","authors":"A. Alvarado, Ricardo Corrales, Maria José Soares Leal, A. Ossa, R. Mora, Manuel Arroyo, Andrea Gomez, Alan Calderon, Jorge L. Arias-Arias","doi":"10.1109/IWOBI.2018.8464219","DOIUrl":"https://doi.org/10.1109/IWOBI.2018.8464219","url":null,"abstract":"In this paper we present a computational model aimed at characterizing the Zika viral infection at a cellular level based on measurements done on viral Dulbecco plaques over time, and describe our current state of progress in the modeling task. So far we have developed an agent-based simulation model of the dispersion of the virus on the cells conforming the viral plaque. The growth rate of the viral plaques and the number of cells counted on each plaque were used to characterize the viral infection in terms of parameters related to the fate of infected cells, such as the probability of a cell infecting its neighboring cells and the probability of an infected cell of dying at any given moment. The model can be used to predict viral plaque growth patterns similar to those observed in the laboratory. Our current efforts focus on optimizing the model parameters to fit the experimental data. Further development of the model includes the description of viral infection kinetics of specific viral strains. Our model has been developed using the agent-based modeling language Netlogo [1].","PeriodicalId":127078,"journal":{"name":"2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127352581","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：