Pub Date : 2015-12-17DOI: 10.1109/BTAS.2015.7358772
A. Morales, Mario Falanga, Julian Fierrez, Carlo Sansone, J. Ortega-Garcia
This work proposes a new benchmark for keystroke dynamics recognition on the basis of fully reproducible research. Instead of traditional authentication approaches based on complex passwords, we propose a novel keystroke recognition based on typing patterns from personal data. We present a new database made up with the keystroke patterns of 63 users and 7560 samples. The proposed approach eliminates the necessity to memorize complex passwords (something that we know) by replacing them by personal data (something that we are). The results encourage to further explore this new application scenario and the availability of data and source code represent a new valuable resource for the research community.
{"title":"Keystroke dynamics recognition based on personal data: A comparative experimental evaluation implementing reproducible research","authors":"A. Morales, Mario Falanga, Julian Fierrez, Carlo Sansone, J. Ortega-Garcia","doi":"10.1109/BTAS.2015.7358772","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358772","url":null,"abstract":"This work proposes a new benchmark for keystroke dynamics recognition on the basis of fully reproducible research. Instead of traditional authentication approaches based on complex passwords, we propose a novel keystroke recognition based on typing patterns from personal data. We present a new database made up with the keystroke patterns of 63 users and 7560 samples. The proposed approach eliminates the necessity to memorize complex passwords (something that we know) by replacing them by personal data (something that we are). The results encourage to further explore this new application scenario and the availability of data and source code represent a new valuable resource for the research community.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129726058","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 : 2015-12-17DOI: 10.1109/BTAS.2015.7358749
Di Miao, Man Zhang, Haiqing Li, Zhenan Sun, T. Tan
Both high accuracy of iris biometrics and friendly interface of face recognition are important issues to a biometric recognition system. So an open problem is how to combine iris and face biometrics for reliable personal identification. This paper proposes a bin-based weak classifier fusion method for Multibiometrics of Iris and Face. The matching scores of iris and face image patches are partitioned into multiple bins so that the weak classifiers are learned on the bins. Such a non-linear score mapping is simple and efficient but it can discover detailed and distinctive information hidden in matching scores. So that pattern classification performance of the matching scores can be significantly improved. In addition, an ensemble learning method based on boosting is used to select the most discriminant and robust bin-based weak classifiers for identity verification. The excellent performance on the CASIA-Iris-Distance demonstrates the advantages of the proposed method over other multibiometric fusion methods.
{"title":"Bin-based weak classifier fusion of iris and face biometrics","authors":"Di Miao, Man Zhang, Haiqing Li, Zhenan Sun, T. Tan","doi":"10.1109/BTAS.2015.7358749","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358749","url":null,"abstract":"Both high accuracy of iris biometrics and friendly interface of face recognition are important issues to a biometric recognition system. So an open problem is how to combine iris and face biometrics for reliable personal identification. This paper proposes a bin-based weak classifier fusion method for Multibiometrics of Iris and Face. The matching scores of iris and face image patches are partitioned into multiple bins so that the weak classifiers are learned on the bins. Such a non-linear score mapping is simple and efficient but it can discover detailed and distinctive information hidden in matching scores. So that pattern classification performance of the matching scores can be significantly improved. In addition, an ensemble learning method based on boosting is used to select the most discriminant and robust bin-based weak classifiers for identity verification. The excellent performance on the CASIA-Iris-Distance demonstrates the advantages of the proposed method over other multibiometric fusion methods.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125843231","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 : 2015-12-17DOI: 10.1109/BTAS.2015.7358769
K. Raja, Ramachandra Raghavendra, C. Busch
Iris recognition in the visible spectrum has many challenging aspects. Especially, for subjects with dark iris color, which is caused by higher melanin pigmentation and collagen fibrils, the pattern is not clearly observable under visible light. Thus, the verification performance is generally lowered due to limited texture visibility in the captured iris samples. In this work, we propose a novel method of employing a white light-emitting-diode (LED) to obtain high-quality iris images with detailed texture. To evaluate the proposed set-up with LED light, we have acquired a new database of dark iris images comprising of 62 unique iris instances with ten samples each that were captured in different sessions. The database is acquired using three different smartphones - iPhone 5S, Nokia Lumia 1020 and Samsung Active S4. We also provide a benchmark of the proposed method with conventional to Near-Infra-Red (NIR) images, which are available for a subset of the database. Extensive experiments were carried out using five different well-established iris recognition algorithms and one commercial-of-the-shelf algorithm. They demonstrate the reliable performance of the proposed image capturing setup with GMR of 91.01% at FMR = 0.01% indicating the applicability in real-life authentication scenarios.
{"title":"Iris imaging in visible spectrum using white LED","authors":"K. Raja, Ramachandra Raghavendra, C. Busch","doi":"10.1109/BTAS.2015.7358769","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358769","url":null,"abstract":"Iris recognition in the visible spectrum has many challenging aspects. Especially, for subjects with dark iris color, which is caused by higher melanin pigmentation and collagen fibrils, the pattern is not clearly observable under visible light. Thus, the verification performance is generally lowered due to limited texture visibility in the captured iris samples. In this work, we propose a novel method of employing a white light-emitting-diode (LED) to obtain high-quality iris images with detailed texture. To evaluate the proposed set-up with LED light, we have acquired a new database of dark iris images comprising of 62 unique iris instances with ten samples each that were captured in different sessions. The database is acquired using three different smartphones - iPhone 5S, Nokia Lumia 1020 and Samsung Active S4. We also provide a benchmark of the proposed method with conventional to Near-Infra-Red (NIR) images, which are available for a subset of the database. Extensive experiments were carried out using five different well-established iris recognition algorithms and one commercial-of-the-shelf algorithm. They demonstrate the reliable performance of the proposed image capturing setup with GMR of 91.01% at FMR = 0.01% indicating the applicability in real-life authentication scenarios.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"115 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115000531","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 : 2015-12-17DOI: 10.1109/BTAS.2015.7358762
Ramachandra Raghavendra, M. Avinash, S. Marcel, C. Busch
Finger vein recognition has emerged as an accurate and reliable biometric modality that was deployed in various security applications. However, the use of finger vein recognition also indicated its vulnerability to presentation attacks (or direct attacks). In this work, we present a novel algorithm to identify the liveness of the finger vein characteristic that is presented to the sensor. The core idea of the proposed approach is to magnify the blood flow through the finger vein to measure its liveness. To this extent, we employ the Eulerian Video Magnification (EVM) approach to enhance the motion of the blood in the recorded finger vein video. Next, we further process the magnified video to extract the motion-based features using optical flow to identify the finger vein artefacts. Extensive experiments are carried out on a relatively large database that is comprised of normal presentations vein videos from 300 unique finger instances corresponding to 100 subjects. The finger vein artefact database is captured by printing 300 real (or normal) presentation image of the finger vein sample on a high-quality paper using two different kinds of printers namely laser and inkjet. Extensive comparative evaluation with four different well-established state-of-the-art schemes demonstrated the efficacy of the proposed scheme.
{"title":"Finger vein liveness detection using motion magnification","authors":"Ramachandra Raghavendra, M. Avinash, S. Marcel, C. Busch","doi":"10.1109/BTAS.2015.7358762","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358762","url":null,"abstract":"Finger vein recognition has emerged as an accurate and reliable biometric modality that was deployed in various security applications. However, the use of finger vein recognition also indicated its vulnerability to presentation attacks (or direct attacks). In this work, we present a novel algorithm to identify the liveness of the finger vein characteristic that is presented to the sensor. The core idea of the proposed approach is to magnify the blood flow through the finger vein to measure its liveness. To this extent, we employ the Eulerian Video Magnification (EVM) approach to enhance the motion of the blood in the recorded finger vein video. Next, we further process the magnified video to extract the motion-based features using optical flow to identify the finger vein artefacts. Extensive experiments are carried out on a relatively large database that is comprised of normal presentations vein videos from 300 unique finger instances corresponding to 100 subjects. The finger vein artefact database is captured by printing 300 real (or normal) presentation image of the finger vein sample on a high-quality paper using two different kinds of printers namely laser and inkjet. Extensive comparative evaluation with four different well-established state-of-the-art schemes demonstrated the efficacy of the proposed scheme.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"535 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128468107","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 : 2015-12-17DOI: 10.1109/BTAS.2015.7358754
Mohammad Rafiqul Alam, Bennamoun, R. Togneri, Ferdous Sohel
This paper presents applications of special types of deep neural networks (DNNs) for audio-visual biometrics. A common example is the DBN-DNN that uses the generative weights of deep belief networks (DBNs) to initialize the feature detecting layers of deterministic feed forward DNNs. In this paper, we propose the DBM-DNN that uses the generative weights of deep Boltzmann machines (DBMs) for initialization of DNNs. Then, a softmax layer is added on top and the DNNs are trained discriminatively. Our experimental results show that lower error rates can be achieved using the DBM-DNN compared to the support vector machine (SVM), linear regression-based classifier (LRC) and the DBN-DNN. Experiments were carried out on two publicly available audio-visual datasets: the VidTIMIT and MOBIO.
{"title":"A deep neural network for audio-visual person recognition","authors":"Mohammad Rafiqul Alam, Bennamoun, R. Togneri, Ferdous Sohel","doi":"10.1109/BTAS.2015.7358754","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358754","url":null,"abstract":"This paper presents applications of special types of deep neural networks (DNNs) for audio-visual biometrics. A common example is the DBN-DNN that uses the generative weights of deep belief networks (DBNs) to initialize the feature detecting layers of deterministic feed forward DNNs. In this paper, we propose the DBM-DNN that uses the generative weights of deep Boltzmann machines (DBMs) for initialization of DNNs. Then, a softmax layer is added on top and the DNNs are trained discriminatively. Our experimental results show that lower error rates can be achieved using the DBM-DNN compared to the support vector machine (SVM), linear regression-based classifier (LRC) and the DBN-DNN. Experiments were carried out on two publicly available audio-visual datasets: the VidTIMIT and MOBIO.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132817899","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 : 2015-12-17DOI: 10.1109/BTAS.2015.7358790
K. Raja, Ramachandra Raghavendra, C. Busch
Biometrics systems are being challenged at the sensor level using artefact presentation such as printed artefacts or electronic screen attacks. In this work, we propose a novel technique to detect the artefact iris images by decomposing the images into Laplacian pyramids of various scales and obtain frequency responses in different orientations. The obtained features are classified using a support vector machine with a polynomial kernel. Further, we extend the same technique with majority voting rule to provide the decision on artefact detection for video based iris recognition in the visible spectrum. The proposed technique is evaluated on the newly created visible spectrum iris video database and also Near-Infra-Red (NIR) images. The newly constructed visible spectrum iris video database is specifically tailored to study the vulnerability of presentation attacks on visible spectrum iris recognition using videos on a smartphone. The newly constructed database is referred as `Presentation Attack Video Iris Database' (PAVID) and consists of 152 unique iris patterns obtained from two different smartphone - iPhone 5S and Nokia Lumia 1020. The proposed technique has provided an Attack Classificiation Error Rate (ACER) of 0.64% on PAVID database and 1.37% on LiveDet iris dataset validating the robustness and applicability of the proposed presentation attack detection (PAD) algorithm in real life scenarios.
{"title":"Presentation attack detection using Laplacian decomposed frequency response for visible spectrum and Near-Infra-Red iris systems","authors":"K. Raja, Ramachandra Raghavendra, C. Busch","doi":"10.1109/BTAS.2015.7358790","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358790","url":null,"abstract":"Biometrics systems are being challenged at the sensor level using artefact presentation such as printed artefacts or electronic screen attacks. In this work, we propose a novel technique to detect the artefact iris images by decomposing the images into Laplacian pyramids of various scales and obtain frequency responses in different orientations. The obtained features are classified using a support vector machine with a polynomial kernel. Further, we extend the same technique with majority voting rule to provide the decision on artefact detection for video based iris recognition in the visible spectrum. The proposed technique is evaluated on the newly created visible spectrum iris video database and also Near-Infra-Red (NIR) images. The newly constructed visible spectrum iris video database is specifically tailored to study the vulnerability of presentation attacks on visible spectrum iris recognition using videos on a smartphone. The newly constructed database is referred as `Presentation Attack Video Iris Database' (PAVID) and consists of 152 unique iris patterns obtained from two different smartphone - iPhone 5S and Nokia Lumia 1020. The proposed technique has provided an Attack Classificiation Error Rate (ACER) of 0.64% on PAVID database and 1.37% on LiveDet iris dataset validating the robustness and applicability of the proposed presentation attack detection (PAD) algorithm in real life scenarios.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"114 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133281580","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 : 2015-12-17DOI: 10.1109/BTAS.2015.7358787
Felix Juefei-Xu, M. Savvides
We propose a new method called the Pokerface for extreme face illumination normalization. The Pokerface is a two-phase approach. It first aims at maximizing the minimum gap between adjacently-valued pixels while keeping the partial ordering of the pixels in the face image under extreme illumination condition, an intuitive effort based on order theory to unveil the underlying structure of a dark image. This optimization can be formulated as a feasibility search problem and can be efficiently solved by linear programming. It then smooths the intermediate representation by repressing the energy of the gradient map. The smoothing step is carried out by total variation minimization and sparse approximation. The illumination normalized faces using our proposed Pokerface not only exhibit very high fidelity against neutrally illuminated face, but also allow for a significant improvement in face verification experiments using even the simplest classifier. Simultaneously achieving high level of faithfulness and expressiveness is very rare among other methods. These conclusions are drawn after benchmarking our algorithm against 22 prevailing illumination normalization techniques on both the CMU Multi-PIE database and Extended YaleB database that are widely adopted for face illumination problems.
{"title":"Pokerface: Partial order keeping and energy repressing method for extreme face illumination normalization","authors":"Felix Juefei-Xu, M. Savvides","doi":"10.1109/BTAS.2015.7358787","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358787","url":null,"abstract":"We propose a new method called the Pokerface for extreme face illumination normalization. The Pokerface is a two-phase approach. It first aims at maximizing the minimum gap between adjacently-valued pixels while keeping the partial ordering of the pixels in the face image under extreme illumination condition, an intuitive effort based on order theory to unveil the underlying structure of a dark image. This optimization can be formulated as a feasibility search problem and can be efficiently solved by linear programming. It then smooths the intermediate representation by repressing the energy of the gradient map. The smoothing step is carried out by total variation minimization and sparse approximation. The illumination normalized faces using our proposed Pokerface not only exhibit very high fidelity against neutrally illuminated face, but also allow for a significant improvement in face verification experiments using even the simplest classifier. Simultaneously achieving high level of faithfulness and expressiveness is very rare among other methods. These conclusions are drawn after benchmarking our algorithm against 22 prevailing illumination normalization techniques on both the CMU Multi-PIE database and Extended YaleB database that are widely adopted for face illumination problems.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133592764","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 : 2015-12-17DOI: 10.1109/BTAS.2015.7358783
Serife Seda Kucur Ergunay, E. Khoury, Alexandros Lazaridis, S. Marcel
Automatic speaker verification (ASV) systems are subject to various kinds of malicious attacks. Replay, voice conversion and speech synthesis attacks drastically degrade the performance of a standard ASV system by increasing its false acceptance rates. This issue raised a high level of interest in the speech research community where the possible voice spoofing attacks and their related countermeasures have been investigated. However, much less effort has been devoted in creating realistic and diverse spoofing attack databases that foster researchers to correctly evaluate their countermeasures against attacks. The existing studies are not complete in terms of types of attacks, and often difficult to reproduce because of unavailability of public databases. In this paper we introduce the voice spoofing data-set of AVspoof, a public audio-visual spoofing database. AVspoof includes ten realistic spoofing threats generated using replay, speech synthesis and voice conversion. In addition, we provide a set of experimental results that show the effect of such attacks on current state-of-the-art ASV systems.
{"title":"On the vulnerability of speaker verification to realistic voice spoofing","authors":"Serife Seda Kucur Ergunay, E. Khoury, Alexandros Lazaridis, S. Marcel","doi":"10.1109/BTAS.2015.7358783","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358783","url":null,"abstract":"Automatic speaker verification (ASV) systems are subject to various kinds of malicious attacks. Replay, voice conversion and speech synthesis attacks drastically degrade the performance of a standard ASV system by increasing its false acceptance rates. This issue raised a high level of interest in the speech research community where the possible voice spoofing attacks and their related countermeasures have been investigated. However, much less effort has been devoted in creating realistic and diverse spoofing attack databases that foster researchers to correctly evaluate their countermeasures against attacks. The existing studies are not complete in terms of types of attacks, and often difficult to reproduce because of unavailability of public databases. In this paper we introduce the voice spoofing data-set of AVspoof, a public audio-visual spoofing database. AVspoof includes ten realistic spoofing threats generated using replay, speech synthesis and voice conversion. In addition, we provide a set of experimental results that show the effect of such attacks on current state-of-the-art ASV systems.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"107 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130694688","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 : 2015-12-17DOI: 10.1109/BTAS.2015.7358799
Yuhang Wu, Xiang Xu, S. Shah, I. Kakadiaris
Head pose estimation helps to align a 3D face model to a 2D image, which is critical to research requiring dense 2D-to-2D or 3D-to-2D correspondence. Traditional pose estimation relies strongly on the accuracy of landmarks, so it is sensitive to missing or incorrect landmarks. In this paper, we propose a landmark-free approach to estimate the pose projection matrix. The method can be used to estimate this matrix in unconstrained scenarios and we demonstrate its effectiveness through multiple head pose estimation experiments.
{"title":"Towards fitting a 3D dense facial model to a 2D image: A landmark-free approach","authors":"Yuhang Wu, Xiang Xu, S. Shah, I. Kakadiaris","doi":"10.1109/BTAS.2015.7358799","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358799","url":null,"abstract":"Head pose estimation helps to align a 3D face model to a 2D image, which is critical to research requiring dense 2D-to-2D or 3D-to-2D correspondence. Traditional pose estimation relies strongly on the accuracy of landmarks, so it is sensitive to missing or incorrect landmarks. In this paper, we propose a landmark-free approach to estimate the pose projection matrix. The method can be used to estimate this matrix in unconstrained scenarios and we demonstrate its effectiveness through multiple head pose estimation experiments.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"30 14","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114116027","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 : 2015-12-17DOI: 10.1109/BTAS.2015.7358765
P. Phillips, Matthew Q. Hill, Jake A. Swindle, A. O’Toole
Face recognition by machines has improved substantially in the past decade and now is at a level that compares favorably with humans for frontal faces acquired by digital single lens reflex cameras. We expand the comparison between humans and algorithms to still images and videos taken with digital point and shoot cameras. The data used for this comparison are from the Point and Shoot Face Recognition Challenge (PaSC). For videos, human performance was compared with the four top performers in the Face and Gesture 2015 Person Recognition Evaluation. In the literature, there are two methods for computing human performance: aggregation and fusion. We show that the fusion method produces higher performance estimates. We report performance for two levels of difficulty: challenging and extremely-difficult. Our results provide additional evidence that human performance shines relative to algorithms on extremely-difficult comparisons. To improve the community's understanding of the state of human and algorithm performance, we update the cross-modal performance analysis in Phillips and O'Toole [22] with these new results.
在过去的十年里，机器的面部识别已经有了很大的进步，现在已经达到了与人类相比的水平，可以通过数码单镜头反光相机获得正面的面部。我们将人与算法之间的比较扩展到用数码相机拍摄的静态图像和视频。用于比较的数据来自Point and Shoot Face Recognition Challenge (PaSC)。对于视频，人类的表现与2015年人脸和手势识别评估中的前四名表现最好的人进行了比较。在文献中，有两种计算人类表现的方法:聚合和融合。我们表明，融合方法产生更高的性能估计。我们根据两个难度等级来报告游戏表现:具有挑战性的和极难的。我们的研究结果提供了额外的证据，证明在极其困难的比较中，人类的表现相对于算法更出色。为了提高社区对人和算法性能状态的理解，我们用这些新结果更新了Phillips和O'Toole[22]的跨模态性能分析。
{"title":"Human and algorithm performance on the PaSC face Recognition Challenge","authors":"P. Phillips, Matthew Q. Hill, Jake A. Swindle, A. O’Toole","doi":"10.1109/BTAS.2015.7358765","DOIUrl":"https://doi.org/10.1109/BTAS.2015.7358765","url":null,"abstract":"Face recognition by machines has improved substantially in the past decade and now is at a level that compares favorably with humans for frontal faces acquired by digital single lens reflex cameras. We expand the comparison between humans and algorithms to still images and videos taken with digital point and shoot cameras. The data used for this comparison are from the Point and Shoot Face Recognition Challenge (PaSC). For videos, human performance was compared with the four top performers in the Face and Gesture 2015 Person Recognition Evaluation. In the literature, there are two methods for computing human performance: aggregation and fusion. We show that the fusion method produces higher performance estimates. We report performance for two levels of difficulty: challenging and extremely-difficult. Our results provide additional evidence that human performance shines relative to algorithms on extremely-difficult comparisons. To improve the community's understanding of the state of human and algorithm performance, we update the cross-modal performance analysis in Phillips and O'Toole [22] with these new results.","PeriodicalId":404972,"journal":{"name":"2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS)","volume":"98 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124887130","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：