Abstract— In this paper, we propose a high performance algorithm for detecting human faces in a still image. Human faces help to communicate and interact in a better way that may be either in human-human interaction or human-machine interaction. The success points for our algorithm is the use of different individual algorithms along with the Unscented Kalman filter (UKF) process, as a novelty of our process. We have modified the algorithms as well. We used Viola Jones eye detector, skin color detector and the Haar cascade classifier for face detection process. Finally, we have conducted a benchmark test for our proposed algorithm using image databases of CMU-MIT, MIT training sets, INRIA Graz-01, and FDDB database. Then, we clarify its effectiveness using ROC curves. 1. INTRODUCTION Nowadays, the face detection process has become very common in computer vision systems. The process of locating the position of the face in an image is known as face detection. The process is performed either by using still or video images. In this paper, we are using still images for face detection. There seems to be a lot of applications and systems applied for face detection, but developing a system that will increase the face detection rate based on the static position of the face on an image will help to make the face detection system more advanced and crucial. Many novel methods have been proposed for face detection. A detailed review over works in the face detection can be found in Ref.1 and 2. Yang [1] provided an approach combining multiple color models for stable color based face detection. Several researches have been done regarding the success of Viola Jones algorithm [3] which describes about the Haar feature calculation for face detection. The process of detecting and aligning faces by image retrieval process and landmark localization [4] shows the face detection process performed under different environmental conditions of occlusion, facial appearances, and pose estimation. The LAB features [5] have been developed from the inspiration of Haar feature and local binary pattern [6] for face detection, these process hold a good result for face detection but still lacks the highest face detection rate. Erdem et al. [7] combined the Haar cascade classifier and skin color detector, but were limited in these two processes only. Luo [8] proposed a face tracking system using modified Viola-Jones method combined with the Kalman filter. However, these studies are applications of the Kalman filter . for tracking. We are developing a new algorithm by using the different face detection algorithms in a different manner. The skin color detector [9] and the Haar cascade classifier are combined with the eye detector [10] for detecting faces and eyes in an image. The skin color detector is slightly modified by adding the low pass filter for removing noise from an image. The Sobel edge detector operates the edges of a face. Moreover, by selecting the skin color region of the face a facial candidate is detected. The detected facial candidate is then passed to an eye detector for checking the presence of eyes in a face. We have modified Haar cascade classifier by combining with the clustering algorithm. It even helps in increasing the accuracy of the face detection rate. After combining these major face detection algorithms, we obtained a good face detection rate, but still some high frequency Gaussian noises were found in the images. So to remove the noises from a detected image, we use the unscented Kalman filter (UKF) process [12], which remove the noise from the images at first and then passes the noise filtered images to the Modified Haar cascade classifier for detecting and verifying the face in an image under the different environmental conditions such as pose, scale, the absence of the structural component, facial expression, occlusion, Illumination variation, color region, multiple face detection and so on. Our applied process is slightly different from the other face detection algorithms implemented yet. The changes, modifications and the differences between our proposed algorithm and the others are mentioned in this paper clearly. As for evaluation metrics, we use the Receiver Operating Characteristic (ROC) curve. Finally, we clarify the effectiveness of our proposed algorithm through the benchmarking process using image databases of CMU-MIT [13], INRIA Graz-01 [14], MIT training sets [15], and FDDB datasets [16]. 2. PROPOSED ALGORITHM FOR FACE DETECTION The main concept of our algorithm is to develop a high performance algorithm for face detection. The skin color detector, Haar cascade classifier, Viola Jones eye detector and the unscented Kalman filter process are used in our proposed algorithm. 2.1 Structure of the proposed algorithm The flow of the proposed algorithm is shown in Figure 1. Firstly, the algorithm is structured in such a way that the skin color detector is modified by using a low pass filter, Sobel edge detector and modified Viola Jones eye detector. Facial candidates and two eyes are detected in this process. Proposing a high performance face detector based on UKF Bikash Lamsal, and Naofumi Matsumoto Correspondence to: Bikash Lamsal E-mail:bikashaitjp@gmail.com Ashikaga Institute of Technology 268-1, Ohmae-cho, Ashikaga, Tochigi 326-8558, Japan Proceedings of the 2014 IEEE/SICE International Symposium on System Integration, Chuo University, Tokyo, Japan, December 13-15, 2014 SaP1E.5 978-1-4799-6943-2/14/$31.00 ©2014 IEEE 298