|
Original research REAL-TIME EMOTION RECOGNITION USING WEARABLE DEVICES AND DEEP LEARNINGPages 95-100
Payal Bute, Akanksha Kadam, Ankita Mehla, Anisha Panwar,
Abstract
This work offers a novel method for recognizing emotions in real time with wearable technology and deep learning algorithms. Utilizing a range of wearable sensors, such as ECG and EEG, the system continuously records physiological data that represent emotional states. Using long short-term memory networks for temporal analysis and convolutional neural networks for feature extraction, the system employs a hybrid deep learning architecture. When compared to conventional single-sensor systems, our approach improves the accuracy of emotion classification. Because of its real-time processing power, the system can be used for a variety of applications, including human-computer interfaces, adaptive entertainment systems, and mental health monitoring. Initial assessments demonstrate its resilience and adaptability in a range of application contexts. Compared to single-modality approaches, the multi-modal approach better represents the complexity of human emotions. Additionally, the study tackles issues such as data unreliability, the study also tackles issues including noise, customized models, and inconsistent data. The system's effectiveness in real-time applications is demonstrated by experimental findings, suggesting applications in healthcare, entertainment, and human-computer interaction.
Keywords: Computer Vision, Deep Learning, Convolution Neural Networks, Real Motion.
|
Harsha Patil 
PDF 