Pub Date : 2018-06-01DOI: 10.1109/IWBF.2018.8401554
Philipp Terhörst, N. Damer, Andreas Braun, Arjan Kuijper
Since fingerprints are one of the most widely deployed biometrics, several applications can benefit from an accurate fingerprint gender estimation. Previous work mainly tackled the task of gender estimation based on complete fingerprints. However, partial fingerprint captures are frequently occurring in many applications including forensics and consumer electronics, with the considered ratio of the fingerprint is variable. Therefore, this work investigates gender estimation on a small, detectable, and well-defined partition of a fingerprint. It investigates gender estimation on the level of a single minutia. Working on this level, we propose a feature extraction process that is able to deal with the rotation and translation invariance problems of fingerprints. This is evaluated on a publicly available database and with five different binary classifiers. As a result, the information of a single minutia achieves a comparable accuracy on the gender classification task as previous work using quarters of aligned fingerprints with an average of more than 25 minutiae.
{"title":"What can a single minutia tell about gender?","authors":"Philipp Terhörst, N. Damer, Andreas Braun, Arjan Kuijper","doi":"10.1109/IWBF.2018.8401554","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401554","url":null,"abstract":"Since fingerprints are one of the most widely deployed biometrics, several applications can benefit from an accurate fingerprint gender estimation. Previous work mainly tackled the task of gender estimation based on complete fingerprints. However, partial fingerprint captures are frequently occurring in many applications including forensics and consumer electronics, with the considered ratio of the fingerprint is variable. Therefore, this work investigates gender estimation on a small, detectable, and well-defined partition of a fingerprint. It investigates gender estimation on the level of a single minutia. Working on this level, we propose a feature extraction process that is able to deal with the rotation and translation invariance problems of fingerprints. This is evaluated on a publicly available database and with five different binary classifiers. As a result, the information of a single minutia achieves a comparable accuracy on the gender classification task as previous work using quarters of aligned fingerprints with an average of more than 25 minutiae.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128539263","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 : 2018-06-01DOI: 10.1109/IWBF.2018.8401565
Ondrej Kanich, M. Drahanský, M. Mézl
The aim of this article is to describe the usage of creative materials for fingerprint spoofs. The majority of these materials are used in some kind of modeling (hence creative materials). In total 21 materials were tested. PCB mold was created from fingerprints taken by various semi-cooperative methods. Using this mold the first set of materials was tested. From these materials the best ones were chosen to be the second stage. Spoofs from the second stage were acquired by several sensors and evaluated by NFIQ and COTS product software. The most promising materials which were tested are latex, Siligum and wax sheet.
{"title":"Use of creative materials for fingerprint spoofs","authors":"Ondrej Kanich, M. Drahanský, M. Mézl","doi":"10.1109/IWBF.2018.8401565","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401565","url":null,"abstract":"The aim of this article is to describe the usage of creative materials for fingerprint spoofs. The majority of these materials are used in some kind of modeling (hence creative materials). In total 21 materials were tested. PCB mold was created from fingerprints taken by various semi-cooperative methods. Using this mold the first set of materials was tested. From these materials the best ones were chosen to be the second stage. Spoofs from the second stage were acquired by several sensors and evaluated by NFIQ and COTS product software. The most promising materials which were tested are latex, Siligum and wax sheet.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129821231","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 : 2018-06-01DOI: 10.1109/IWBF.2018.8401568
Dogucan Yaman, Fevziye Irem Eyiokur, N. Sezgin, H. K. Ekenel
In this paper, we present a detailed analysis on extracting soft biometrie traits, age and gender, from ear images. Although there have been a few previous work on gender classification using ear images, to the best of our knowledge, this study is the first work on age classification from ear images. In the study, we have utilized both geometric features and appearance-based features for ear representation. The utilized geometric features are based on eight anthropometric landmarks and consist of 14 distance measurements and two area calculations. The appearance-based methods employ deep convolutional neural networks for representation and classification. The well-known convolutional neural network models, namely, AlexNet, VGG-16, GoogLeNet, and SqueezeNet have been adopted for the study. They have been fine-tuned on a large-scale ear dataset that has been built from the profile and close-to-profile face images in the Multi-PIE face dataset. This way, we have performed a domain adaptation. The updated models have been fine-tuned once more time on the small-scale target ear dataset, which contains only around 270 ear images for training. According to the experimental results, appearance-based methods have been found to be superior to the methods based on geometric features. We have achieved 94% accuracy for gender classification, whereas 52% accuracy has been obtained for age classification. These results indicate that ear images provide useful cues for age and gender classification, however, further work is required for age estimation.
{"title":"Age and gender classification from ear images","authors":"Dogucan Yaman, Fevziye Irem Eyiokur, N. Sezgin, H. K. Ekenel","doi":"10.1109/IWBF.2018.8401568","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401568","url":null,"abstract":"In this paper, we present a detailed analysis on extracting soft biometrie traits, age and gender, from ear images. Although there have been a few previous work on gender classification using ear images, to the best of our knowledge, this study is the first work on age classification from ear images. In the study, we have utilized both geometric features and appearance-based features for ear representation. The utilized geometric features are based on eight anthropometric landmarks and consist of 14 distance measurements and two area calculations. The appearance-based methods employ deep convolutional neural networks for representation and classification. The well-known convolutional neural network models, namely, AlexNet, VGG-16, GoogLeNet, and SqueezeNet have been adopted for the study. They have been fine-tuned on a large-scale ear dataset that has been built from the profile and close-to-profile face images in the Multi-PIE face dataset. This way, we have performed a domain adaptation. The updated models have been fine-tuned once more time on the small-scale target ear dataset, which contains only around 270 ear images for training. According to the experimental results, appearance-based methods have been found to be superior to the methods based on geometric features. We have achieved 94% accuracy for gender classification, whereas 52% accuracy has been obtained for age classification. These results indicate that ear images provide useful cues for age and gender classification, however, further work is required for age estimation.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121372921","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 : 2018-06-01DOI: 10.1109/IWBF.2018.8401562
U. Scherhag, C. Rathgeb, C. Busch
In past years, different researchers have shown the vulnerability of face recognition systems to attacks based on morphed face images. More recently, first morph detection subsystems have been proposed to automatically detect this kind of fraud. While some algorithms have been reported to reveal practical detection performance on individual datasets a systematic analysis of proposed detectors with respect to their robustness across different databases has remained elusive. In this work, we evaluate the performance of different morph detection algorithms across disjoint datasets of 2,745 bona fide and 14,337 automatically generated morphed face images. Within a generic evaluation framework a systematic robustness estimation scheme is proposed to identify reliable detection algorithms. Finally, the robustness of algorithms which have been determined as most promising is verified on another disjoint dataset. Hence, this paper represents the first attempt towards a comprehensive cross-database performance evaluation and a systematic evaluation of the robustness of morphed face image detection algorithms.
{"title":"Performance variation of morphed face image detection algorithms across different datasets","authors":"U. Scherhag, C. Rathgeb, C. Busch","doi":"10.1109/IWBF.2018.8401562","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401562","url":null,"abstract":"In past years, different researchers have shown the vulnerability of face recognition systems to attacks based on morphed face images. More recently, first morph detection subsystems have been proposed to automatically detect this kind of fraud. While some algorithms have been reported to reveal practical detection performance on individual datasets a systematic analysis of proposed detectors with respect to their robustness across different databases has remained elusive. In this work, we evaluate the performance of different morph detection algorithms across disjoint datasets of 2,745 bona fide and 14,337 automatically generated morphed face images. Within a generic evaluation framework a systematic robustness estimation scheme is proposed to identify reliable detection algorithms. Finally, the robustness of algorithms which have been determined as most promising is verified on another disjoint dataset. Hence, this paper represents the first attempt towards a comprehensive cross-database performance evaluation and a systematic evaluation of the robustness of morphed face image detection algorithms.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129695003","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 : 2018-06-01DOI: 10.1109/IWBF.2018.8401555
L. Debiasi, U. Scherhag, C. Rathgeb, A. Uhl, C. Busch
In the recent past, face recognition systems have been found to be highly vulnerable to attacks based on morphed biometrie samples. Such attacks pose a severe security threat to biometric recognition systems across various applications. Apart from some algorithms, which have been reported to reveal practical detection performance on small in-house datasets, approaches to effectively detect morphed face images of high quality have remained elusive. In this paper, we propose a morph detection algorithm based on an analysis of photo response non-uniformity (PRNU). It is based on a spectral analysis of the variations within the PRNU caused by the morphing process. On a comprehensive database of 961 bona fide and 2,414 morphed face images practical performance in terms of detection equal error rate (D-EER) is achieved. Additionally, the robustness of the proposed morph detection algorithm towards different post-processing procedures, e.g. histogram equalization or sharpening, is assessed.
{"title":"PRNU-based detection of morphed face images","authors":"L. Debiasi, U. Scherhag, C. Rathgeb, A. Uhl, C. Busch","doi":"10.1109/IWBF.2018.8401555","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401555","url":null,"abstract":"In the recent past, face recognition systems have been found to be highly vulnerable to attacks based on morphed biometrie samples. Such attacks pose a severe security threat to biometric recognition systems across various applications. Apart from some algorithms, which have been reported to reveal practical detection performance on small in-house datasets, approaches to effectively detect morphed face images of high quality have remained elusive. In this paper, we propose a morph detection algorithm based on an analysis of photo response non-uniformity (PRNU). It is based on a spectral analysis of the variations within the PRNU caused by the morphing process. On a comprehensive database of 961 bona fide and 2,414 morphed face images practical performance in terms of detection equal error rate (D-EER) is achieved. Additionally, the robustness of the proposed morph detection algorithm towards different post-processing procedures, e.g. histogram equalization or sharpening, is assessed.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129999121","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 : 2018-06-01DOI: 10.1109/IWBF.2018.8401556
G. Koukiou, V. Anastassopoulos
Fusion of dissimilar features by means of neural networks is demonstrated in this work aiming at improving the performance of these features for drunk person identification. The features are coming from the thermal images of the face of the inspected persons and have been derived using different image analysis techniques. Thus, they convey dissimilar information, which has to be transferred onto the same framework and fused to result into a decision with improved reliability. Conventional data association techniques are employed to explore the available information. After that, fusion of the information is carried out using Neural Networks. The resulting decision is of higher reliability compared to those achieved using the individual features separately. Experimental results are provided based on an existing sober-drunk database. The main advantage of the method is that it is not invasive and all the information is acquired remotely. In practice, an electronic system incorporating the proposed approach will point out to the police to whom an extended inspection for alcohol consumption is due.
{"title":"Fusion using neural networks for intoxication identification","authors":"G. Koukiou, V. Anastassopoulos","doi":"10.1109/IWBF.2018.8401556","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401556","url":null,"abstract":"Fusion of dissimilar features by means of neural networks is demonstrated in this work aiming at improving the performance of these features for drunk person identification. The features are coming from the thermal images of the face of the inspected persons and have been derived using different image analysis techniques. Thus, they convey dissimilar information, which has to be transferred onto the same framework and fused to result into a decision with improved reliability. Conventional data association techniques are employed to explore the available information. After that, fusion of the information is carried out using Neural Networks. The resulting decision is of higher reliability compared to those achieved using the individual features separately. Experimental results are provided based on an existing sober-drunk database. The main advantage of the method is that it is not invasive and all the information is acquired remotely. In practice, an electronic system incorporating the proposed approach will point out to the police to whom an extended inspection for alcohol consumption is due.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128553967","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 : 2018-06-01DOI: 10.1109/IWBF.2018.8401560
Sudipta Banerjee, A. Ross
The principle of Photo Response Non Uniformity (PRNU) is often used to link a digital image with the sensor that produced it. In this regard, a number of schemes have been proposed in the literature to extract PRNU details from a given input image. In this work, we study the impact of photometric transformations applied to near-infrared ocular images, on PRNU-based iris sensor identification accuracy. The contributions of this work are as follows: (a) Firstly, we evaluate the impact of 7 different photometric transformations on 4 different PRNU-based sensor identification schemes; (b) Secondly, we develop an explanatory model based on the Jensen-Shannon divergence measure to analyze the conditions under which these PRNU estimation schemes fail on photometrically transformed images. The analysis is conducted using 9,626 ocular images pertaining to 11 different iris sensors. Experiments suggest that (a) the Enhanced Sensor Pattern Noise and Maximum Likelihood Estimation based Sensor Pattern Noise techniques are more robust to photometric transformations than other PRNU-based schemes; (b) the application of photometric transformations actually improves the performance of the Phase Sensor Pattern Noise scheme; (c) the single-scale Self Quotient Image (SQI) and Difference of Gaussians (DoG) filtering transformations adversely impact all 4 PRNU-based schemes considered in this work; and (d) the Jensen-Shannon divergence measure is able to explain the degradation in performance of PRNU-based schemes as a function of the photometrically modified images.
{"title":"Impact of photometric transformations on PRNU estimation schemes: A case study using near infrared ocular images","authors":"Sudipta Banerjee, A. Ross","doi":"10.1109/IWBF.2018.8401560","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401560","url":null,"abstract":"The principle of Photo Response Non Uniformity (PRNU) is often used to link a digital image with the sensor that produced it. In this regard, a number of schemes have been proposed in the literature to extract PRNU details from a given input image. In this work, we study the impact of photometric transformations applied to near-infrared ocular images, on PRNU-based iris sensor identification accuracy. The contributions of this work are as follows: (a) Firstly, we evaluate the impact of 7 different photometric transformations on 4 different PRNU-based sensor identification schemes; (b) Secondly, we develop an explanatory model based on the Jensen-Shannon divergence measure to analyze the conditions under which these PRNU estimation schemes fail on photometrically transformed images. The analysis is conducted using 9,626 ocular images pertaining to 11 different iris sensors. Experiments suggest that (a) the Enhanced Sensor Pattern Noise and Maximum Likelihood Estimation based Sensor Pattern Noise techniques are more robust to photometric transformations than other PRNU-based schemes; (b) the application of photometric transformations actually improves the performance of the Phase Sensor Pattern Noise scheme; (c) the single-scale Self Quotient Image (SQI) and Difference of Gaussians (DoG) filtering transformations adversely impact all 4 PRNU-based schemes considered in this work; and (d) the Jensen-Shannon divergence measure is able to explain the degradation in performance of PRNU-based schemes as a function of the photometrically modified images.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133406823","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 : 2018-06-01DOI: 10.1109/IWBF.2018.8401550
O. Nikisins, Teodors Eglitis, André Anjos, S. Marcel
Most of the research on vein biometrics addresses the problems of either palm or finger vein recognition with a considerably smaller emphasis on wrist vein modality. This paper paves the way to a better understanding of capabilities and challenges in the field of wrist vein verification. This is achieved by introducing and discussing a fully automatic cross-correlation based wrist vein verification technique. Overcoming the limitations of ordinary cross-correlation, the proposed system is capable of compensating for scale, translation and rotation between vein patterns in a computationally efficient way. Introduced comparison algorithm requires only two cross-correlation operations to compensate for both translation and rotation, moreover the well known property of log-polar transformation of Fourier magnitudes is not involved in any form. To emphasize the veins, a two-layer Hessian-based vein enhancement approach with adaptive brightness normalization is introduced, improving the connectivity and the stability of extracted vein patterns. The experiments on the publicly available PUT Vein wrist database give promising results with FNMR of 3.75% for FMR « 0.1%. In addition we make this research reproducible providing the source code and instructions to replicate all findings in this work.
{"title":"Fast cross-correlation based wrist vein recognition algorithm with rotation and translation compensation","authors":"O. Nikisins, Teodors Eglitis, André Anjos, S. Marcel","doi":"10.1109/IWBF.2018.8401550","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401550","url":null,"abstract":"Most of the research on vein biometrics addresses the problems of either palm or finger vein recognition with a considerably smaller emphasis on wrist vein modality. This paper paves the way to a better understanding of capabilities and challenges in the field of wrist vein verification. This is achieved by introducing and discussing a fully automatic cross-correlation based wrist vein verification technique. Overcoming the limitations of ordinary cross-correlation, the proposed system is capable of compensating for scale, translation and rotation between vein patterns in a computationally efficient way. Introduced comparison algorithm requires only two cross-correlation operations to compensate for both translation and rotation, moreover the well known property of log-polar transformation of Fourier magnitudes is not involved in any form. To emphasize the veins, a two-layer Hessian-based vein enhancement approach with adaptive brightness normalization is introduced, improving the connectivity and the stability of extracted vein patterns. The experiments on the publicly available PUT Vein wrist database give promising results with FNMR of 3.75% for FMR « 0.1%. In addition we make this research reproducible providing the source code and instructions to replicate all findings in this work.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134643054","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 : 2018-06-01DOI: 10.1109/IWBF.2018.8401558
Mateusz Trokielewicz, A. Czajka
This paper presents a method for segmenting iris images obtained from the deceased subjects, by training a deep convolutional neural network (DCNN) designed for the purpose of semantic segmentation. Post-mortem iris recognition has recently emerged as an alternative, or additional, method useful in forensic analysis. At the same time it poses many new challenges from the technological standpoint, one of them being the image segmentation stage, which has proven difficult to be reliably executed by conventional iris recognition methods. Our approach is based on the SegNet architecture, fine-tuned with 1,300 manually segmented post-mortem iris images taken from the Warsaw-BioBase-Post-Mortem-Iris v1.0 database. The experiments presented in this paper show that this data-driven solution is able to learn specific deformations present in post-mortem samples, which are missing from alive irises, and offers a considerable improvement over the state-of-the-art, conventional segmentation algorithm (OSIRIS): the Intersection over Union (IoU) metric was improved from 73.6% (for OSIRIS) to 83% (for DCNN-based presented in this paper) averaged over subject-disjoint, multiple splits of the data into train and test subsets. This paper offers the first known to us method of automatic processing of post-mortem iris images. We offer source codes with the trained DCNN that perform end-to-end segmentation of post-mortem iris images, as described in this paper. Also, we offer binary masks corresponding to manual segmentation of samples from Warsaw-BioBase-Post-Mortem-Iris v1.0 database to facilitate development of alternative methods for post-mortem iris segmentation.
{"title":"Data-driven segmentation of post-mortem iris images","authors":"Mateusz Trokielewicz, A. Czajka","doi":"10.1109/IWBF.2018.8401558","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401558","url":null,"abstract":"This paper presents a method for segmenting iris images obtained from the deceased subjects, by training a deep convolutional neural network (DCNN) designed for the purpose of semantic segmentation. Post-mortem iris recognition has recently emerged as an alternative, or additional, method useful in forensic analysis. At the same time it poses many new challenges from the technological standpoint, one of them being the image segmentation stage, which has proven difficult to be reliably executed by conventional iris recognition methods. Our approach is based on the SegNet architecture, fine-tuned with 1,300 manually segmented post-mortem iris images taken from the Warsaw-BioBase-Post-Mortem-Iris v1.0 database. The experiments presented in this paper show that this data-driven solution is able to learn specific deformations present in post-mortem samples, which are missing from alive irises, and offers a considerable improvement over the state-of-the-art, conventional segmentation algorithm (OSIRIS): the Intersection over Union (IoU) metric was improved from 73.6% (for OSIRIS) to 83% (for DCNN-based presented in this paper) averaged over subject-disjoint, multiple splits of the data into train and test subsets. This paper offers the first known to us method of automatic processing of post-mortem iris images. We offer source codes with the trained DCNN that perform end-to-end segmentation of post-mortem iris images, as described in this paper. Also, we offer binary masks corresponding to manual segmentation of samples from Warsaw-BioBase-Post-Mortem-Iris v1.0 database to facilitate development of alternative methods for post-mortem iris segmentation.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130910901","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 : 2018-06-01DOI: 10.1109/IWBF.2018.8401559
Souad Khellat-Kihel, A. Lagorio, M. Tistarelli
Face recognition has a strong potential for identity verification on mobile devices, now embedding high resolution cameras and high-end computing hardware. Personal computing devices often also embed automatic face detection, thus facilitating the extraction and processing of face data. The main objective of this paper is to implement a flexible architecture to recognize faces from partial face data. The proposed architecture can be very effective to analyze video data from forensic cases where portions of the face are hidden from other objects. The proposed approach is based on the application of Kernel Fisher Analysis (KFA) to Gabor features extracted from the available face data. Several experiments carried out on realistic image samples demonstrate the validity of the proposed approach.
{"title":"Face recognition “on the move” combining incomplete information","authors":"Souad Khellat-Kihel, A. Lagorio, M. Tistarelli","doi":"10.1109/IWBF.2018.8401559","DOIUrl":"https://doi.org/10.1109/IWBF.2018.8401559","url":null,"abstract":"Face recognition has a strong potential for identity verification on mobile devices, now embedding high resolution cameras and high-end computing hardware. Personal computing devices often also embed automatic face detection, thus facilitating the extraction and processing of face data. The main objective of this paper is to implement a flexible architecture to recognize faces from partial face data. The proposed architecture can be very effective to analyze video data from forensic cases where portions of the face are hidden from other objects. The proposed approach is based on the application of Kernel Fisher Analysis (KFA) to Gabor features extracted from the available face data. Several experiments carried out on realistic image samples demonstrate the validity of the proposed approach.","PeriodicalId":259849,"journal":{"name":"2018 International Workshop on Biometrics and Forensics (IWBF)","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132102151","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：