Context-free attentional operators: the generalized symmetry transform
International Journal of Computer Vision - Special issue on qualitative vision
Efficient illumination normalization of facial images
Pattern Recognition Letters
Example-Based Learning for View-Based Human Face Detection
IEEE Transactions on Pattern Analysis and Machine Intelligence
Face Detection in Color Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
Digital Image Processing
Journal of Cognitive Neuroscience
Face detection using quantized skin color regions merging andwavelet packet analysis
IEEE Transactions on Multimedia
Automatic image segmentation by integrating color-edge extraction and seeded region growing
IEEE Transactions on Image Processing
Face segmentation using skin-color map in videophone applications
IEEE Transactions on Circuits and Systems for Video Technology
Hi-index | 0.00 |
We proposed an efficient method for detecting an identifiable face (ID face) that is not veiled in an input image to establish his or her identity. This method is especially designed to be robust to color degradation and facial veiling, which is composed of two parts: face-like region segmentation and unveiled-face detection. In the face-like region segmentation, measuring horizontal symmetry that is an important property of facial components is introduced to overcome the difficulty of facial region segmentation only by using skin color (SC) under nonuniform illumination causing severe color degradation. As a result, the segmentation leads to extraction of non-SC facial components and their neighbor degraded SC regions as well as undegraded SC regions. The unveiled-face detection is based on analysis of face constellations and statistical averages of facial patterns. The detection especially investigates statistical averages of each facial component pattern and its horizontal symmetry, which leads to detection of a face where all facial components are unveiled. Experimental results for AR and VCL facial databases show that the proposed method yields the improvement of 22.9% in detection rate over a face detection method without consideration of color degradation and facial veiling.