Indonesian Journal of Electrical Engineering and Computer Science Vol. 16, No. 2, November 2019, pp. 827∼834 ISSN: 2502-4752, DOI: 10.11591/ijeecs.v16i2.pp827-834 ❒ 827 An optimization of facial feature point detection program by using several types of convolutional neural network Shyota Shindo 1 , Takaaki Goto 2 , Tadaaki Kirishima 3 , Kensei Tsuchida 4 1,3,4 Toyo University, 2100 Kujirai, Kawagoe, Saitama, Japan 2 Ryutsu Keizai University, 3-2-1 Shin-Matsudo, Matsudo, Chiba, Japan Article Info Article history: Received Jan 17, 2019 Revised Apr 7, 2019 Accepted May 10, 2019 Keywords: Facial feature point detection Neural network Convolutional neural network ABSTRACT Detection of facial feature points is an important technique used for biometric au- thentication and facial expression estimation. A facial feature point is a local point indicating both ends of the eye, holes of the nose, and end points of the mouth in the face image. Many researches on face feature point detection have been done so far, but the accuracy of facial organ point detection is improving by the approach using Con- volutional Neural Network (CNN). However, CNN not only takes time to learn but also the neural network becomes a complicated model, so it is necessary to improve learning time and detection accuracy. In this research, the improvement of the detec- tion accuracy of the learning speed is improved by increasing the convolution layer. Copyright c  2019 Insitute of Advanced Engineeering and Science. All rights reserved. Corresponding Author: Takaaki Goto, Ryutsu Keizai University, 3-2-1 Shin-Matsudo, Matsudo, Chiba, Japan. Email: tg@gotolab.net 1. INTRODUCTION A facial feature point is a local point indicating a place such as an eye end or a mouth end of a facial image. The detection of facial feature points is applied to important technologies such as facial expression estimation and biometric authentication using facial images. Many detection methods have been proposed so far, but with the advent of Convolutional Neural Network (CNN) in recent years, many researches on detection methods using CNN have been conducted, and detection with higher accuracy is getting expected [1]. However, CNN learning takes time. If the layer of CNN becomes deep and the number of training data is large, the learning time becomes huge. As a method for speeding up learning, there are methods using GPU with good performance, and methods for devising hardware such as adding main memory. In addition, the methods for devising software are [2, 3, 4]. Among them, there are a few methods [5] to devise pre-processing for input data. In this paper, we aim at improving preprocessing of input data and speed up learning of facial feature point detection program using CNN. CNN was implemented in Python with reference to the program of Ya- mashita et al. [6]. We propose a method to reduce the number of layers of CNN by applying Laplacian filter to preprocessing and reducing image features. 2. RELATED WORKS Facial feature point detection can be obtained by various methods such as CNN and image processing. As a conventional method, Cootes et al’s Active Appearance Model (AAM) is available [7]. In this method, the average Shape is obtained by using the coordinate points of the facial images of all the learning data, Journal homepage: http://iaescore.com/journals/index.php/ijeecs