Facial Expression Recognition Using Local Gravitational Force Descriptor-Based Deep Convolution Neural Networks

Abstract

An image is worth a thousand words; hence, a face image illustrates extensive details about the specification, gender, age, and emotional states of mind. Facial expressions play an important role in community-based interactions and are often used in the behavioral analysis of emotions. Recognition of automatic facial expressions from a facial image is a challenging task in the computer vision community and admits a large set of applications, such as driver safety, human-computer interactions, health care, behavioral science, video conferencing, cognitive science, and others. In this work, a deep-learning-based scheme is proposed for identifying the facial expression of a person. The proposed method consists of two parts. The former one finds out local features from face images using a local gravitational force descriptor, while, in the latter part, the descriptor is fed into a novel deep convolution neural network (DCNN) model. The proposed DCNN has two branches. The first branch explores geometric features, such as edges, curves, and lines, whereas holistic features are extracted by the second branch. Finally, the scorelevel fusion technique is adopted to compute the final classification score. The proposed method along with 25 state-of-the-art methods is implemented on five benchmark available databases,

Facial Expression Recognition Based on Deep Learning Convolution Neural Network: A Review

“…Moreover 6 researchers ( [25], [28], [61], [62], [64] and [65]) relied on DCNN with its built-in extraction and classification features.in [62] the author used maximum pooling method for feature extraction, and softmax as classification method.in [25], they proposed DCNN that has two branches. The first branch examines geometric features, such as edges, curves, and arcs, while the second branch extracts holistic features.…”

“…Next, an image input RGB of m×n size is read and converted into a gray image with the standard equation [24]. The circumference of the face was detected with Haar [25], [26], [27] Cascade pictures library. Those rectangular facial expressions were then cut off and reported to the same scale.…”

“…As a deep extractor for extracting image functionality. These features are then used to train conventional classifications, including Support Vector Machine (SVM), Discriminant Analysis classification (LDA), and K-Nearest Neighbors (K-NN) [25], [37]. In the extraction process, several complicated properties are formed to clarify the expression of facial form or texture modifications.…”

“…Each of the conventional classifiers has its own features. For instance, SVM [25]. To distinguish several data groups, kernels can transform several feature representations into higher dimensional space.…”

“…Mohan et al [25] created a new DCNN system to categorize facial expressions using holistic features of features. In this way, a biological feature extractor is used to retrieve the local low-level features before using the proposed Deep Belief network (DCNN) model.…”