• Ms. Arti Ahirwar
• Mr. Sarvesh Singh Rai,
• Dr. Navdeep Kaur Saluja

In recent years, we've seen a remarkable rise in the use of artificial intelligence (AI) and machine learning across various sectors, especially in the realm of mental health assessment. This project presents a fresh approach called "Artificial Intelligence based Facial Emotions Analysis for Depression Detection." Its goal is to harness AI and deep learning to pinpoint and classify levels of depression by analyzing facial emotions. Developed in the Matlab programming environment, this project employs the AlexNet Convolutional Neural Network (CNN) model to achieve accurate emotion recognition. The primary aim of this research is to create a reliable system that can identify five essential emotions—Anger, Disgust, Happy, Neutral, and Sadness—by analyzing facial expressions in images. These emotions serve as vital indicators for assessing a person's mental state, particularly when it comes to detecting depression. The system not only identifies these emotions but also categorizes depression into four distinct levels: No Depression, Mild Depression, Moderate Depression, and High Depression. This multi-class classification method provides a deeper insight into an individual's mental health status. To ensure high accuracy in both emotion recognition and depression classification, the AlexNet CNN model is employed. Known for its deep architecture and outstanding feature extraction capabilities, this model is an excellent fit for complex image analysis tasks. After a thorough training process using a diverse dataset that includes facial expressions of varying intensities, the system boasts an impressive accuracy rate of 99%.

Artificial intelligence, Machine learning, Facial Expression, Deep leaning, Viola Jones, AlexNet CNN, depression detection.

"An Explainable Deep Learning Framework for Depression Screening using Attention-based CNNs and Facial Landmark Localization ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 1, page no.d498-d506, January-2026, Available :http://www.jetir.org/papers/JETIR2601362.pdf

"An Explainable Deep Learning Framework for Depression Screening using Attention-based CNNs and Facial Landmark Localization ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 1, page no. ppd498-d506, January-2026, Available at : http://www.jetir.org/papers/JETIR2601362.pdf