Frame-Based Facial Expression Recognition Using Geometrical Features

Abstract

To improve the human-computer interaction (HCI) to be as good as human-human interaction, building an efficient approach for human emotion recognition is required. These emotions could be fused from several modalities such as facial expression, hand gesture, acoustic data, and biophysiological data. In this paper, we address the frame-based perception of the universal human facial expressions (happiness, surprise, anger, disgust, fear, and sadness), with the help of several geometrical features. Unlike many other geometry-based approaches, the frame-based method does not rely on prior knowledge of a person-specific neutral expression; this knowledge is gained through human intervention and not available in real scenarios. Additionally, we provide a method to investigate the performance of the geometry-based approaches under various facial point localization errors. From an evaluation on two public benchmark datasets, we have found that using eight facial points, we can achieve the state-of-the-art recognition rate. However, this state-of-the-art geometry-based approach exploits features derived from 68 facial points and requires prior knowledge of the person-specific neutral expression. The expression recognition rate using geometrical features is adversely affected by the errors in the facial point localization, especially for the expressions with subtle facial deformations.

Effective semantic features for facial expressions recognition using SVM

“…They attain an average AU recognition rate of 95.3 % on a benchmark set and 72 % when tested on spontaneous expressions. In addition, Saeed et al [31] use eight facial points to achieve state-of-the-art recognition rate using a SVM. However, the expression recognition rate using geometrical features is adversely affected by the errors in the facial point localization, especially for the expressions with subtle facial deformations.…”

“…4 System architecture According to the previous discussions, most researches [3,9,28,30,31,38,39,41] focus on the classification module of facial expressions without considering face detection module. This may be due to the challenge caused by complex environments, lighting, or face orientations.…”

“…Classification algorithms for facial expression recognition [49] are usually space-based methods like neural networks [25], support vector machine (SVM) [3,9,28,31,39,41], kNearest-Neighbor (k-NN) [31], AdaBoost method [47], etc. Time and space based methods are like hidden Markov model (HMM) [33], regression neural network, spatial and temporal motion energy templates method.…”

Recognition of Facial Expressions Based on Tracking and Selection of Discriminative Geometric Features

“…A coupled scale Gaussian process regression (CSGPR) model is used in order to normalize the head pose to the frontal pose in terms of extracted facial landmarks. Recently, in [6], geometric features based on only eight facial feature points extracted from single frame are used to recognize the facial expression. Ghimire and Lee [7] proposed a method for facial expression recognition using geometric features extracted from salient points and lines composed of facial key points in the temporal domain, which are selected using AdaBoost algorithm.…”

Yüz ifadelerinin otomatik analizi üzerine bir literatür çalışması

“…Bu yaklaşıma dayanan yöntemlerde yüz bileşenleri ya da yüz özellik noktaları, yüz geometrisini temsil eden bir özellik vektörü kullanılarak çıkartılır [63]. Geometrik özelliklerin ölçümü genellikle yüz bölgesinin analizine, özellikle de yüz bölgesindeki önemli noktaları bulunmasına ya da izlemesine bağlıdır ve literatürde bu konuda yapılan çeşitli çalışmalar vardır [64][65][66][67]. Yapılan çalışmalarda Aktif Şekil Modeli (ASM-Active Shape Model) [68,69] ve Optik Akış [70,71] en çok kullanılan yöntemlerdir.…”