In this paper, we propose an approach for detecting facial features and recovering pose in presence of high out of plane rotations for both still images and video streams. To detect the correct features, we assign a confidence number to combinations of feature candidates given the edge map of the face. Feature candidates are determined using probability distribution of color space of skin, eyes and eyebrows.

To increase the accuracy of feature detection for video streams, we incorporate motion history information for individual features by weighing the confidence measure according to potentialregions of features. Once the best feature combination is obtained, we recover the pose using the centroid of the features assuming orthographic projection.
We conducted experiments on both still images and eleven video sequences including two CNN interviews. In most of the cases, the system performed very well and correctly determined the pose.

Extracting face features and recovering pose are two challenging problems in computer vision, which have been widely explored by researchers. Many high-level vision applications such as video telephony, face recognition, facial animation, facial feature tracking and MPEG-4 coding, require feature extraction and pose recovery, which is currently done manually or semiautomatically for limited orientations such as frontal views.

In videophones, information of the subjectís face, i.e. face orientation and facial expressions, is required for achieving high compression ratio. Similarly, creating realistic facial animation from real images requires initial pose estimation and feature detection of the input face photos for conforming generic wire frame structure to images. Also, most face recognition schemes rely on salient features such as eyes, eyebrows and nose, and relationships between them in 2D without recovering

Download the report from the attachment below on Power Scenario Of India