Pub Date : 2009-09-28DOI: 10.1109/BTAS.2009.5339048
B. Sy
This research investigates slice-based architecture for biometrics. A service slice is an aggregation of resources for a specific biometric objective; e.g., speaker verification. Slice-based architecture is attractive as a framework for modeling service-oriented biometric applications. In order for it to be usable, slice-based architecture must adequately address privacy, security, and standard based interoperability. We propose to incorporate secure computation mechanism and BioAPI standard into slice-based architecture. We discuss why secure computation is information-theoretic secure, and how it can be used to realize a private computation for the exchange of biometric data between two parties in slice-based architecture. For proof-of-concept, open source software is developed for realizing a privacy preserving voice verification prototype based on slice-based architecture, and will be released for experimentation by the public. The result of our initial experimentation is reported.
{"title":"Slice-based architecture for biometrics: Prototype illustration on privacy preserving voice verification","authors":"B. Sy","doi":"10.1109/BTAS.2009.5339048","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339048","url":null,"abstract":"This research investigates slice-based architecture for biometrics. A service slice is an aggregation of resources for a specific biometric objective; e.g., speaker verification. Slice-based architecture is attractive as a framework for modeling service-oriented biometric applications. In order for it to be usable, slice-based architecture must adequately address privacy, security, and standard based interoperability. We propose to incorporate secure computation mechanism and BioAPI standard into slice-based architecture. We discuss why secure computation is information-theoretic secure, and how it can be used to realize a private computation for the exchange of biometric data between two parties in slice-based architecture. For proof-of-concept, open source software is developed for realizing a privacy preserving voice verification prototype based on slice-based architecture, and will be released for experimentation by the public. The result of our initial experimentation is reported.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131228897","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 : 2009-09-28DOI: 10.1109/BTAS.2009.5339059
K. Nandakumar, A. Ross, Anil K. Jain
Sources of information in a multibiometric system are often assumed to be statistically independent in order to simplify the design of the fusion algorithm. However, the independence assumption may not be always valid. In this paper, we analyze whether modeling the dependence between match scores in a multibiometric system has any effect on the fusion performance. Our analysis is based on the likelihood ratio (LR) based fusion framework, which guarantees optimal performance if the match score densities are known. We show that the assumption of independence between matchers has a significant negative impact on the performance of the LR fusion scheme only when (i) the dependence characteristics among genuine match scores is different from that of the impostor scores and (ii) the individual matchers are not very accurate.
{"title":"Biometric fusion: Does modeling correlation really matter?","authors":"K. Nandakumar, A. Ross, Anil K. Jain","doi":"10.1109/BTAS.2009.5339059","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339059","url":null,"abstract":"Sources of information in a multibiometric system are often assumed to be statistically independent in order to simplify the design of the fusion algorithm. However, the independence assumption may not be always valid. In this paper, we analyze whether modeling the dependence between match scores in a multibiometric system has any effect on the fusion performance. Our analysis is based on the likelihood ratio (LR) based fusion framework, which guarantees optimal performance if the match score densities are known. We show that the assumption of independence between matchers has a significant negative impact on the performance of the LR fusion scheme only when (i) the dependence characteristics among genuine match scores is different from that of the impostor scores and (ii) the individual matchers are not very accurate.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131422110","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 : 2009-09-28DOI: 10.1109/BTAS.2009.5339074
N. Houmani, S. Garcia-Salicetti, B. Dorizzi
In this work, we study different combinations of the five time functions captured by a digitizer in presence or not of time variability. To this end, we propose two criteria independent of the classification step: Personal Entropy, introduced in our previous works and an intra-class variability measure based on Dynamic Time Warping. We confront both criteria to system performance using a Hidden Markov Model (HMM) and Dynamic Time Warping (DTW). Moreover, we introduce the concept of short-term time variability, proposed on MCYT-100, and long-term time variability studied with BIOMET database. Our experiments clarify conflicting results in the literature and confirm some other: pen inclination angles are very unstable in presence or not of time variability; the only combination which is robust to time variability is that containing only coordinates; finally, pen pressure is not recommended in the long-term context, although it may give better results in terms of performance (according to the classifier used) in the short-term context.
{"title":"On assessing the robustness of pen coordinates, pen pressure and pen inclination to time variability with personal entropy","authors":"N. Houmani, S. Garcia-Salicetti, B. Dorizzi","doi":"10.1109/BTAS.2009.5339074","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339074","url":null,"abstract":"In this work, we study different combinations of the five time functions captured by a digitizer in presence or not of time variability. To this end, we propose two criteria independent of the classification step: Personal Entropy, introduced in our previous works and an intra-class variability measure based on Dynamic Time Warping. We confront both criteria to system performance using a Hidden Markov Model (HMM) and Dynamic Time Warping (DTW). Moreover, we introduce the concept of short-term time variability, proposed on MCYT-100, and long-term time variability studied with BIOMET database. Our experiments clarify conflicting results in the literature and confirm some other: pen inclination angles are very unstable in presence or not of time variability; the only combination which is robust to time variability is that containing only coordinates; finally, pen pressure is not recommended in the long-term context, although it may give better results in terms of performance (according to the classifier used) in the short-term context.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132487788","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 : 2009-09-28DOI: 10.1109/BTAS.2009.5339056
J. Gentile, N. Ratha, J. Connell
There have been claims of very high information content in iris texture, higher even than in fingerprints. This makes iris attractive for large scale identification systems with possibly millions of people. However, some systems operate by performing N 1:1 matches of the probe against the database. This can get prohibitively expensive in terms of computation as N grows large. Note that for identification systems the per-match time dominants system performance, unlike verification where feature extraction time is the primary component. In this paper we show how to use a short-length iris code to pre-screen a large database and thereby reduce the number of full comparisons needed to a fraction of the total. Since the screening code is much smaller than the full iris code, the time to process the whole database is greatly reduced. As an added benefit, we show that we can use the alignment inferred from the short code to greatly restrict the range of alignments searched for the full code, which further speeds up the system. As we demonstrate in experiments, the two stage approach can reduce the cost and/or time needed by an order of magnitude with very little impact on identification performance.
{"title":"An efficient, two-stage iris recognition system","authors":"J. Gentile, N. Ratha, J. Connell","doi":"10.1109/BTAS.2009.5339056","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339056","url":null,"abstract":"There have been claims of very high information content in iris texture, higher even than in fingerprints. This makes iris attractive for large scale identification systems with possibly millions of people. However, some systems operate by performing N 1:1 matches of the probe against the database. This can get prohibitively expensive in terms of computation as N grows large. Note that for identification systems the per-match time dominants system performance, unlike verification where feature extraction time is the primary component. In this paper we show how to use a short-length iris code to pre-screen a large database and thereby reduce the number of full comparisons needed to a fraction of the total. Since the screening code is much smaller than the full iris code, the time to process the whole database is greatly reduced. As an added benefit, we show that we can use the alignment inferred from the short code to greatly restrict the range of alignments searched for the full code, which further speeds up the system. As we demonstrate in experiments, the two stage approach can reduce the cost and/or time needed by an order of magnitude with very little impact on identification performance.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133221278","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 : 2009-09-28DOI: 10.1109/BTAS.2009.5339044
G. Amayeh, G. Bebis, M. Nicolescu
Since the biometric data tends to have a large intra-class variability, it is possible for the enrolled templates to be significantly different from acquired samples during system's operation. The majority of existing techniques in the literature, namely self update, update a template set by using a confidently verified input sample in order to avoid the introduction of impostors into the template set of a client. Therefore these techniques can only exploit the input sample very similar to the current template set leading to local optimization of a template set. To address this issue, this paper introduces a technique by decomposing the hand silhouette into the different parts (i.e. fingers) and analyzing the confidences of these parts in order to lead to global optimization of templates. In the proposed method, first the hand silhouette is divided in different parts corresponding to the fingers. Then the confidence of each finger, as well as its identity, is evaluated by a Support Vector Data Description (SVDD). The confidence of a query hand is determined by the maximum confidence of all fingers. If the maximum confidence is higher than a threshold, the boundaries of all fingers' SVDDs are incrementally updated to learn the variations of the input data. The motivation behind this technique is that the temporal changes that may occur in the fingers are uncorrelated in such a way that the confidence of each finger can be significantly different from the others. As a result those fingers with difficult intra-class variations (low confidence) can be used in the update process by this technique. The experimental results show the effectiveness of the proposed technique in comparison to the state of the art self-update technique specially at low false acceptance rates.
{"title":"Improving hand-based verification through online finger template update based on fused confidences","authors":"G. Amayeh, G. Bebis, M. Nicolescu","doi":"10.1109/BTAS.2009.5339044","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339044","url":null,"abstract":"Since the biometric data tends to have a large intra-class variability, it is possible for the enrolled templates to be significantly different from acquired samples during system's operation. The majority of existing techniques in the literature, namely self update, update a template set by using a confidently verified input sample in order to avoid the introduction of impostors into the template set of a client. Therefore these techniques can only exploit the input sample very similar to the current template set leading to local optimization of a template set. To address this issue, this paper introduces a technique by decomposing the hand silhouette into the different parts (i.e. fingers) and analyzing the confidences of these parts in order to lead to global optimization of templates. In the proposed method, first the hand silhouette is divided in different parts corresponding to the fingers. Then the confidence of each finger, as well as its identity, is evaluated by a Support Vector Data Description (SVDD). The confidence of a query hand is determined by the maximum confidence of all fingers. If the maximum confidence is higher than a threshold, the boundaries of all fingers' SVDDs are incrementally updated to learn the variations of the input data. The motivation behind this technique is that the temporal changes that may occur in the fingers are uncorrelated in such a way that the confidence of each finger can be significantly different from the others. As a result those fingers with difficult intra-class variations (low confidence) can be used in the update process by this technique. The experimental results show the effectiveness of the proposed technique in comparison to the state of the art self-update technique specially at low false acceptance rates.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133319593","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 : 2009-09-28DOI: 10.1109/BTAS.2009.5339076
H. K. Ekenel, R. Stiefelhagen
In this paper, we present a face registration approach, in which alignment is done by minimizing the closest distance at the classification step. This method eliminates the need of a feature localization step that exists in traditional face recognition systems and formulates alignment as an optimization process during classification. In other words, instead of performing a separate facial feature localization step and localizing facial features according to some type of feature matching score, in the proposed method, alignment is done by directly optimizing the classification score. Moreover, a feature detector can still be integrated to the system. In this case, the output of the feature detector is used as the initial point of the optimization process. Results of extensive experiments have shown that the proposed approach leads very high correct recognition rates, especially in the case of partial face occlusion, where it is not possible to precisely detect the facial feature locations. It has been also found that, in the case of using a facial feature detector, the approach can tolerate localization errors of up to 18% of the interocular distance.
{"title":"Face alignment by minimizing the closest classification distance","authors":"H. K. Ekenel, R. Stiefelhagen","doi":"10.1109/BTAS.2009.5339076","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339076","url":null,"abstract":"In this paper, we present a face registration approach, in which alignment is done by minimizing the closest distance at the classification step. This method eliminates the need of a feature localization step that exists in traditional face recognition systems and formulates alignment as an optimization process during classification. In other words, instead of performing a separate facial feature localization step and localizing facial features according to some type of feature matching score, in the proposed method, alignment is done by directly optimizing the classification score. Moreover, a feature detector can still be integrated to the system. In this case, the output of the feature detector is used as the initial point of the optimization process. Results of extensive experiments have shown that the proposed approach leads very high correct recognition rates, especially in the case of partial face occlusion, where it is not possible to precisely detect the facial feature locations. It has been also found that, in the case of using a facial feature detector, the approach can tolerate localization errors of up to 18% of the interocular distance.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116176675","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 : 2009-09-28DOI: 10.1109/BTAS.2009.5339058
Bian Yang, C. Busch
In this paper a parameterized geometric alignment method is proposed for minutiae-based fingerprint template protection by transforming an original minutia vicinity into a geometrically-aligned and protected minutia vicinity by randomly generated parameters. Template diversification can be achieved by setting different parameters for different minutiae vicinities. Comparison result of two protected templates is summarized from comparison results of protected minutiae vicinities from both templates. Experimental results on the public FVC2002DB2_A database show satisfactory biometric performance (with average Equal Error Rate 0.0404) of the proposed algorithm. Performance and security analysis are also given for the proposed approach.
{"title":"Parameterized geometric alignment for minutiae-based fingerprint template protection","authors":"Bian Yang, C. Busch","doi":"10.1109/BTAS.2009.5339058","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339058","url":null,"abstract":"In this paper a parameterized geometric alignment method is proposed for minutiae-based fingerprint template protection by transforming an original minutia vicinity into a geometrically-aligned and protected minutia vicinity by randomly generated parameters. Template diversification can be achieved by setting different parameters for different minutiae vicinities. Comparison result of two protected templates is summarized from comparison results of protected minutiae vicinities from both templates. Experimental results on the public FVC2002DB2_A database show satisfactory biometric performance (with average Equal Error Rate 0.0404) of the proposed algorithm. Performance and security analysis are also given for the proposed approach.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127814348","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 : 2009-09-28DOI: 10.1109/BTAS.2009.5339041
Ngoc-Son Vu, A. Caplier
The performance of face recognition systems can be dramatically degraded when the pose of the probe face is different from the gallery face. In this paper, we present a pose robust face recognition model, centered on modeling how face patches change in appearance as the viewpoint varies. We present a novel model based on two robust local appearance descriptors, Gabor wavelets and Local Binary Patterns (LBP). These two descriptors have been widely exploited for face recognition and different strategies for combining them have been investigated. However, to the best of our knowledge, all existing combination methods are designed for frontal face recognition. We introduce a local statistical framework for face recognition across pose variations, given only one frontal reference image. The method is evaluated on the Feret pose dataset and experimental results show that we achieve very high recognition rates over the wide range of pose variations presented in this challenging dataset.
{"title":"Efficient statistical face recognition across pose using Local Binary Patterns and Gabor wavelets","authors":"Ngoc-Son Vu, A. Caplier","doi":"10.1109/BTAS.2009.5339041","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339041","url":null,"abstract":"The performance of face recognition systems can be dramatically degraded when the pose of the probe face is different from the gallery face. In this paper, we present a pose robust face recognition model, centered on modeling how face patches change in appearance as the viewpoint varies. We present a novel model based on two robust local appearance descriptors, Gabor wavelets and Local Binary Patterns (LBP). These two descriptors have been widely exploited for face recognition and different strategies for combining them have been investigated. However, to the best of our knowledge, all existing combination methods are designed for frontal face recognition. We introduce a local statistical framework for face recognition across pose variations, given only one frontal reference image. The method is evaluated on the Feret pose dataset and experimental results show that we achieve very high recognition rates over the wide range of pose variations presented in this challenging dataset.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123501770","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 : 2009-09-28DOI: 10.1109/BTAS.2009.5339027
J. Gentile, N. Ratha, J. Connell
The texture in a human iris has been shown to have good individual distinctiveness and thus is suitable for use in reliable identification. A conventional iris recognition system unwraps the iris image and generates a binary feature vector by quantizing the response of selected filters applied to the rows of this image. Typically there are 360 angular sectors, 64 radial rings, and 2 filter responses. This produces a full-length iris code (FLIC) of about 5760 bytes. In contrast, this paper seeks to shrink the representation by finding those regions of the iris that contain the most descriptive potential. We show through experiments that the regions close to the pupil and sclera contribute least to discrimination, and that there is a high correlation between adjacent radial rings. Using these observations we produce a short-length iris code (SLIC) of only 450 bytes. The SLIC is an order of magnitude smaller the FLIC and yet has comparable performance as shown by results on the MMU2 database. The smaller sized representation has the advantage of being easier to store as a barcode, and also reduces the matching time per pair.
{"title":"SLIC: Short-length iris codes","authors":"J. Gentile, N. Ratha, J. Connell","doi":"10.1109/BTAS.2009.5339027","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339027","url":null,"abstract":"The texture in a human iris has been shown to have good individual distinctiveness and thus is suitable for use in reliable identification. A conventional iris recognition system unwraps the iris image and generates a binary feature vector by quantizing the response of selected filters applied to the rows of this image. Typically there are 360 angular sectors, 64 radial rings, and 2 filter responses. This produces a full-length iris code (FLIC) of about 5760 bytes. In contrast, this paper seeks to shrink the representation by finding those regions of the iris that contain the most descriptive potential. We show through experiments that the regions close to the pupil and sclera contribute least to discrimination, and that there is a high correlation between adjacent radial rings. Using these observations we produce a short-length iris code (SLIC) of only 450 bytes. The SLIC is an order of magnitude smaller the FLIC and yet has comparable performance as shown by results on the MMU2 database. The smaller sized representation has the advantage of being easier to store as a barcode, and also reduces the matching time per pair.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125952783","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 : 2009-09-28DOI: 10.1109/BTAS.2009.5339025
Y. Lui, D. Bolme, B. Draper, J. Beveridge, G. Givens, P. Phillips
This paper presents a meta-analysis for covariates that affect performance of face recognition algorithms. Our review of the literature found six covariates for which multiple studies reported effects on face recognition performance. These are: age of the person, elapsed time between images, gender of the person, the person's expression, the resolution of the face images, and the race of the person. The results presented are drawn from 25 studies conducted over the past 12 years. There is near complete agreement between all of the studies that older people are easier to recognize than younger people, and recognition performance begins to degrade when images are taken more than a year apart. While individual studies find men or women easier to recognize, there is no consistent gender effect. There is universal agreement that changing expression hurts recognition performance. If forced to compare different expressions, there is still insufficient evidence to conclude that any particular expression is better than another. Higher resolution images improve performance for many modern algorithms. Finally, given the studies summarized here, no clear conclusions can be drawn about whether one racial group is harder or easier to recognize than another.
{"title":"A meta-analysis of face recognition covariates","authors":"Y. Lui, D. Bolme, B. Draper, J. Beveridge, G. Givens, P. Phillips","doi":"10.1109/BTAS.2009.5339025","DOIUrl":"https://doi.org/10.1109/BTAS.2009.5339025","url":null,"abstract":"This paper presents a meta-analysis for covariates that affect performance of face recognition algorithms. Our review of the literature found six covariates for which multiple studies reported effects on face recognition performance. These are: age of the person, elapsed time between images, gender of the person, the person's expression, the resolution of the face images, and the race of the person. The results presented are drawn from 25 studies conducted over the past 12 years. There is near complete agreement between all of the studies that older people are easier to recognize than younger people, and recognition performance begins to degrade when images are taken more than a year apart. While individual studies find men or women easier to recognize, there is no consistent gender effect. There is universal agreement that changing expression hurts recognition performance. If forced to compare different expressions, there is still insufficient evidence to conclude that any particular expression is better than another. Higher resolution images improve performance for many modern algorithms. Finally, given the studies summarized here, no clear conclusions can be drawn about whether one racial group is harder or easier to recognize than another.","PeriodicalId":325900,"journal":{"name":"2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2009-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125865727","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：