Electronics, Journal Year: 2025, Volume and Issue: 14(7), P. 1320 - 1320

Published: March 27, 2025

This study advances artificial intelligence by pioneering the classification of human emotions (for patients) with a healthcare mobile robot, anticipating human–robot interaction for humans (patients) admitted in hospitals or any environment. delves into challenge accurately classifying emotion as patient emotion, which is critical factor understanding patients’ recent moods and situations. We integrate convolutional neural networks (CNNs), recurrent (RNNs), multi-layer perceptrons (MLPs) to analyze facial comprehensively. The process begins deploying faster region-based network (Faster R-CNN) swiftly identify real-time recorded video feeds. includes advanced feature extraction across three CNN models innovative fusion techniques, strengthen improved Inception-V3 superior accuracy replace Faster R-CNN learning module. valuable replacement aims enhance face detection our proposed framework. Carefully acquired these datasets simulated Validation on EMOTIC, CK+, FER-2013, AffectNet all showed impressive rates 98.01%, 99.53%, 99.27%, 96.81%, respectively. These class-wise show that it has potential advance medical environment measures intelligent manufacturing robots.

Language: Английский