Automatic face image annotation using machine learning techniques

Face image annotation refers to automatically assign labels to objects in a face image. Assigned labels may be gender, age, name, etc. For annotating the human faces, wide variety of algorithms and techniques are used in computer vision. In this paper, a real time face image annotation system which involves face detection, gender and age identification is proposed. For detection of human faces in an image Viola and Jones Object detection technique is used. For the detected human faces, features are extracted using the process of Local Binary Pattern. For classification of human faces based on their gender and age, Support Vector Machine is used by considering extracted features as support vectors. The main application of this work is to count the number of human faces who have viewed a particular advertisement. The face database consists of sample face dataset of varying age group and gender. The experimental results show better accuracy for gender and age classification.