[PDF][PDF] Improved Reliable Deep Face Recognition Method Using Separated Components
H Chamani, M Nadjafi - Journal of Brilliant Engineering, 2022 - acapublishing.com
H Chamani, M Nadjafi
Journal of Brilliant Engineering, 2022•acapublishing.com… We also intend to examine methods other than point averaging in more detail so that we
may be able to prevent the loss of isolated component information. We intend to use our
method, DCSFR, in addition to the current datasets, on the MEGA Face database, which
has about 5 million images of about 672,000 identities, and examine the result because we
believe that the deep convolution network can achieve better results with providing a large
number of samples, even with the DCSFR method. …
may be able to prevent the loss of isolated component information. We intend to use our
method, DCSFR, in addition to the current datasets, on the MEGA Face database, which
has about 5 million images of about 672,000 identities, and examine the result because we
believe that the deep convolution network can achieve better results with providing a large
number of samples, even with the DCSFR method. …
Hossein Chamani1, Mohammad Nadjafi*, 2 1Master of Computer Engineering, Tehran, Iran 2Assistance Professor of Aerospace Engineering, Aerospace Research Institute (Ministry of Science, Research and Technology), Tehran, Iran
acapublishing.com
以上显示的是最相近的搜索结果。 查看全部搜索结果