Illumination normalization based on correction of large-scale components for face recognition

2017 ◽  
Vol 266 ◽  
pp. 465-476 ◽  
Author(s):  
Xiaoguang Tu ◽  
Jingjing Gao ◽  
Mei Xie ◽  
Jin Qi ◽  
Zheng Ma
2017 ◽  
Vol 9 (3) ◽  
pp. 334-339
Author(s):  
Rokas Semėnas

Face recognition programs have many practical usages in various fields, such as security or entertainment. Existing recognition algorithms must deal with various real life problems – mainly with illumination. In practice, illumination normalization models are often used only for Small-scale futures extraction, ignoring Large-scale features. In this article, new and more direct approach to this problem is offered, used algorithms and test results are given.


2013 ◽  
Vol 32 (7) ◽  
pp. 2049-2052
Author(s):  
Chao-you LI ◽  
Ji-zhou SUN
Keyword(s):  

2021 ◽  
Author(s):  
◽  
Ella Macaskill

<p>Face recognition is a fundamental cognitive function that is essential for social interaction – yet not everyone has it. Developmental prosopagnosia is a lifelong condition in which people have severe difficulty recognising faces but have normal intellect and no brain damage. Despite much research, the component processes of face recognition that are impaired in developmental prosopagnosia are not well understood. Two core processes are face perception, being the formation of visual representations of a currently seen face, and face memory, being the storage, maintenance, and retrieval of those representations. Most studies of developmental prosopagnosia focus on face memory deficits, but a few recent studies indicate that face perception deficits might also be important. Characterising face perception in developmental prosopagnosia is crucial for a better understanding of the condition. In this thesis, I addressed this issue in a large-scale experiment with 108 developmental prosopagnosics and 136 matched controls. I assessed face perception abilities with multiple measures and ran a broad range of analyses to establish the severity, scope, and nature of face perception deficits in developmental prosopagnosia. Three major results stand out. First, face perception deficits in developmental prosopagnosia were severe, and could be comparable in size to face memory deficits. Second, the face perception deficits were widespread, affecting the whole sample rather than a subset of individuals. Third, the deficits were mainly driven by impairments to mechanisms specialised for processing upright faces. Further analyses revealed several other features of the deficits, including the use of atypical and inconsistent strategies for perceiving faces, difficulties matching the same face across different pictures, equivalent impact of lighting and viewpoint variations in face images, and atypical perceptual and non-perceptual components of test performance. Overall, my thesis shows that face perception deficits are more central to developmental prosopagnosia than previously thought and motivates further research on the issue.</p>


Author(s):  
Yandong Guo ◽  
Lei Zhang ◽  
Yuxiao Hu ◽  
Xiaodong He ◽  
Jianfeng Gao
Keyword(s):  

Sign in / Sign up

Export Citation Format

Share Document