Отрывок: This transformation maps raw classifier score to a [0, 1] interval. The correspondence we used be- tween error rates and mapped score values is presented in Table 2. Here FNR stands for False Negative Rate, and FPR stands for False Positive Rate. In our experiments we used additional held-out dataset for scores mapping weights computation. Table 2. Correspondence used for score mapping estimation Key point Mapped score FNR = 0.0001 0.1 ...
Полная запись метаданных
Поле DC | Значение | Язык |
---|---|---|
dc.contributor.author | Nikitin, M.Yu. | - |
dc.contributor.author | Konushin, V.S. | - |
dc.contributor.author | Konushin, A.S. | - |
dc.date.accessioned | 2019-10-15 10:03:46 | - |
dc.date.available | 2019-10-15 10:03:46 | - |
dc.date.issued | 2019-08 | - |
dc.identifier | Dspace\SGAU\20190924\78787 | ru |
dc.identifier.citation | Nikitin MYu, Konushin VS, Konushin AS. Face anti-spoofing with joint spoofing medium detection and eye blinking analysis. Computer Optics 2019; 43(4): 618-626. DOI: 10.18287/2412-6179-2019-43-4-618-626. | ru |
dc.identifier.uri | https://dx.doi.org/10.18287/2412-6179-2019-43-4-618-626 | - |
dc.identifier.uri | http://repo.ssau.ru/handle/Zhurnal-Komputernaya-optika/Face-antispoofing-with-joint-spoofing-medium-detection-and-eye-blinking-analysis-78787 | - |
dc.description.abstract | Modern biometric systems based on face recognition demonstrate high recognition quality, but they are vulnerable to face presentation attacks, such as photo or replay attack. Existing face anti-spoofing methods are mostly based on texture analysis and due to lack of training data either use hand-crafted features or fine-tuned pretrained deep models. In this paper we present a novel CNN-based approach for face anti-spoofing, based on joint analysis of the presence of a spoofing medium and eye blinking. For training our classifiers we propose the procedure of synthetic data generation which allows us to train powerful deep models from scratch. Experimental analysis on the challenging datasets (CASIA-FASD, NUUA Imposter) shows that our method can obtain state-of-the-art results. | ru |
dc.language.iso | en_US | ru |
dc.publisher | Новая техника | ru |
dc.relation.ispartofseries | 43;4 | - |
dc.subject | face anti-spoofing | ru |
dc.subject | synthetic data | ru |
dc.subject | video analysis | ru |
dc.subject | neural networks | ru |
dc.subject | deep learning | ru |
dc.title | Face anti-spoofing with joint spoofing medium detection and eye blinking analysis | ru |
dc.type | Article | ru |
dc.textpart | This transformation maps raw classifier score to a [0, 1] interval. The correspondence we used be- tween error rates and mapped score values is presented in Table 2. Here FNR stands for False Negative Rate, and FPR stands for False Positive Rate. In our experiments we used additional held-out dataset for scores mapping weights computation. Table 2. Correspondence used for score mapping estimation Key point Mapped score FNR = 0.0001 0.1 ... | - |
Располагается в коллекциях: | Журнал "Компьютерная оптика" |
Файлы этого ресурса:
Файл | Описание | Размер | Формат | |
---|---|---|---|---|
430414.pdf | Основная статья | 1.14 MB | Adobe PDF | Просмотреть/Открыть |
Показать базовое описание ресурса
Просмотр статистики
Поделиться:
Все ресурсы в архиве электронных ресурсов защищены авторским правом, все права сохранены.