Role of Zoning in Facial Expression Using Deep Learning

Abstract

Facial expression is an unspoken message essential to collaboration and effective discourse. An inner emotional state of a human is expressed using facial expressions and is very effective for communication with actual emotions. Anger, happiness, sadness, contempt, surprise, fear, disgust, and neutral are eight common expressions of humans. Scientific community proposed several face emotion recognition techniques. However, due to fewer face landmarks and their intensity for deep learning models, performance improvement for facial expression recognition still needs to be improved for accurately predicting facial emotion recognition. This study proposes a zoning-based face expression recognition (ZFER) to locate more face landmarks to perceive deep face emotions indemnity through zoning. After face extraction, landmarks from the face, such as the eyes, eyebrows, nose, forehead, and mouth, are extracted. The second step is zoning each landmark into four regions and zone-based face landmarks are passed to the VGG-16 model to generate a feature map. Finally, the feature map is given as input to fully connected neural network (FCNN) to classify facial emotions into multiple classes. Various experiments are performed on facial expression recognition (FER) 2013 and CK+ datasets to evaluate our proposed model with state-ofthe-art facial expression recognition approaches using performance assessment metrics like accuracy. The accuracy of the proposed method with face features on CK+ and FER2013 are 98.4% and 65%, respectively. The experimental zoning results improve from 98.47% to 98.74% on the CK+ dataset.

Facial emotion recognition based on auxiliary classifiers and SENet module

“…Compared to other models, the proposed model achieves a 68.87% emotion recognition accuracy with only 0.46 million parameters. Methods Parameter(M) Accuracy(%) MobileNetV3(Large) [9] 4.2 63.9 Efficient-SwishNet [10] 5.31 64.2 CNN [11] 0.93 65.77 Shahzad et al [12] -…”

Context Transformer and Adaptive Method with Visual Transformer for Robust Facial Expression Recognition

“…Results on CK+: Table 7 shows that we achieved a high recognition performance on the CK+ dataset by using a 10-fold cross-validation method with 99.20% accuracy. Among them, FER_RN [49] introduces an attention module to redistribute the weight parameters of channel and spatial dimensions, and ZFER [50] performs emotion recognition on faces based on partitions. These methods only take into account local feature information, while our model pairs have a more comprehensive understanding of expression information through the CNN_ViT architecture.…”

“…SL+SSLpuzzling [51] 2021 98.23 FER_RN [49] 2022 96.97 CFNet [31] 2023 99.07 DBN [32] 2023 98.19 CNN_LSTM [52] 2023 92.00 ZFER [50] 2023 98.74 CoT_AdaptiveViT(Ours) 2024 99.20…”

Lightweight EfficientNetB3 Model Based on Depthwise Separable Convolutions for Enhancing Classification of Leukemia White Blood Cell Images

“…ALL blast cells and normal cells are difficult to identify because of their similarities. The CNN technique is one of the most advanced and popular computer vision techniques to efficiently utilize for different tasks related to processing image data [9], [10]. Various medical imaging applications successfully used pre-trained neural networks like ResNet, VGGnet, and Inception.…”