Related Experiment Video
Updated: Sep 14, 2025
06:37
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
Published on: December 15, 2023
4.1K
An interactive information based DCNN-BiLSTM model with dual attention mechanism for facial expression recognition
Samanthisvaran Jayaraman1, Anand Mahendran2
1School of Computer Science and Engineering, Vellore Institute of Technology, Vellore, Tamilnadu, India.
Scientific Reports
|July 19, 2025
Summary
This study introduces a novel Deep Convolutional Neural Network with Bi-Long Short Term Memory and attention mechanisms for accurate facial expression recognition. The model achieves superior performance across multiple datasets, demonstrating its effectiveness in classifying human emotions.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Machine Learning
Background:
- Human facial expressions significantly influence actions and decisions.
- Basic Convolutional Neural Networks (CNNs) face challenges in operational speed and complexity.
- Accurate facial expression recognition is crucial for human-computer interaction and behavioral analysis.
Purpose of the Study:
- To develop an efficient and accurate model for facial expression recognition.
- To improve the classification of human emotions from facial images.
- To address the complexity and speed limitations of existing CNN models.
Main Methods:
- Proposed a Deep Convolutional Neural Network (DCNN) integrated with Bi-Long Short Term Memory (Bi-LSTM).
- Incorporated a single and cross-fusion attention mechanism for spatial and channel information extraction.
- Utilized Piecewise Cubic Polynomial and linear activation functions for faster Interactive Learning Information (ILI) processing.
- Employed Global Average Pooling (GAP) and a softmax classifier for emotion classification into 7 categories.
Main Results:
- Achieved high accuracy rates: 82.89% (FER 2013), 96.78% (CK+), 95.78% (RAF-DB), and 95.87% (JAFFE).
- Demonstrated lower False Recognition Rates (FAR): 7.23% (FER 2013), 1.42% (CK+), 1.96% (RAF-DB), and 1.78% (JAFFE).
- Exhibited superior Genuine Recognition Rates (GAR): 88.57% (FER2013), 97.23% (CK+), 96.87% (RAF-DB), and 96.32% (JAFFE).
Conclusions:
- The proposed DCNN with Bi-LSTM and attention mechanism outperforms existing benchmarking methods.
- The model offers a significant advancement in facial expression recognition accuracy and efficiency.
- This approach holds promise for real-world applications requiring reliable emotion detection.
Related Concept Videos
01:21Association Areas of the Cortex
6.3K
Association areas are regions of the cerebral cortex that do not have a specific sensory or motor function. Instead, they integrate and interpret information from various sources to enable higher cognitive processes such as memory, learning, and decision-making. Some key association areas include the following:
Prefrontal Association Area: This area is located in the frontal lobe and is involved in planning, decision-making, and moderating social behavior. It connects with primary motor areas,...
Prefrontal Association Area: This area is located in the frontal lobe and is involved in planning, decision-making, and moderating social behavior. It connects with primary motor areas,...
6.3K
01:24Facial Feedback Hypothesis
251
Charles Darwin proposed that facial expressions are an evolutionary adaptation for communication. He argued that these expressions are not influenced by culture but are universal across species. For example, a snarling expression with exposed teeth signals a threat in many animals, including humans. Darwin also suggested that displaying an emotion can intensify the feeling. Smiling, for example, could enhance one's sense of happiness. This idea laid the foundation for understanding the role...
251