Fast Face Detection Using Graphics Processor
Fast face detection is one of the key components of various computer vision applications. Viola-Jones algorithm provides a good and fast detection for low and medium resolution images. This paper proposes a new and fast approach to perform real time face detection. The proposed method includes the enhanced Haar-like features and uses SVM for training and classification. Experimental result shows that the authors' proposed architecture works well for high resolution images. Their design model combines conventional programming using CPU with GP-GPU programming using NVIDIA CUDA for fast face detection.