This paper addresses the critical need for accurate stress detection, emphasizing its far-reaching effects on health and social dynamics. It proposes an efficient deep learning (DL) approach for stress detection from facial videos, focusing on remote photoplethysmography (rPPG). The study employs hybrid DL networks, including Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), and 1D Convolutional Neural Network (1D-CNN) models, with hyperparameter optimization and augmentation techniques. The proposed approach achieves up to 95.83% accuracy with the UBFC-Phys dataset while maintaining excellent computational efficiency. The experimental results demonstrate the effectiveness of the proposed Hybrid DL models for rPPG-based stress detection, offering a substantial improvement in accuracy and efficiency. The paper also discusses the methodology, dataset, and performance evaluation, highlighting the advantages and limitations of the proposed approach.This paper addresses the critical need for accurate stress detection, emphasizing its far-reaching effects on health and social dynamics. It proposes an efficient deep learning (DL) approach for stress detection from facial videos, focusing on remote photoplethysmography (rPPG). The study employs hybrid DL networks, including Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), and 1D Convolutional Neural Network (1D-CNN) models, with hyperparameter optimization and augmentation techniques. The proposed approach achieves up to 95.83% accuracy with the UBFC-Phys dataset while maintaining excellent computational efficiency. The experimental results demonstrate the effectiveness of the proposed Hybrid DL models for rPPG-based stress detection, offering a substantial improvement in accuracy and efficiency. The paper also discusses the methodology, dataset, and performance evaluation, highlighting the advantages and limitations of the proposed approach.