• G.V.Vinod
• V.Rakesh
• D.PavanKumar
• V.Mahesh
• V.VarunDev

In this work, In order to solve the face anti-spoofing problem, we propose a novel framework based on the Convolutional Neural Network (CNN) and the Recurrent Neural Network (RNN), which is motivated by the way humans decide whether a given face example is real or fake: by quickly scanning the entire face to get a sense of its overall appearance and then examining its local features in detail to pick out subtle differences (RNN). To be more precise, we apply deep reinforcement learning to characterise the action of exploring face-spoofing-related information in image sub-patches. We also create a recurrent approach to gradually train RNN representations of local information from the investigated sub-patches. In the end, we use a convolutional neural network (CNN) to learn both the local and global features of the input image, and then combine them to make a classification decision. We also conduct extensive experiments on a variety of open-source datasets, including an ablation study and a visualisation analysis, to further evaluate our suggested technique. The results of our experiments reveal that our method is superior to the state-of-the-art since it achieves state-of-the-art performance in every scenario tested.

FACE ANTI-SPOOFING CNN, rPPG, RNN, SVM.

"FACE ANTI-SPOOFING BY COMBINING HIGH- AND LOW-FREQUENCY CHARACTERISTICS FOR SUPERIOR GENERALISATION POWER", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.10, Issue 3, page no.c698-c701, March-2023, Available :http://www.jetir.org/papers/JETIR2303278.pdf

"FACE ANTI-SPOOFING BY COMBINING HIGH- AND LOW-FREQUENCY CHARACTERISTICS FOR SUPERIOR GENERALISATION POWER", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.10, Issue 3, page no. ppc698-c701, March-2023, Available at : http://www.jetir.org/papers/JETIR2303278.pdf