Detection of Alcohol Drug Consumption Among Youngsters using Eye Gazing Movement and Deep Learning Techniques

The intake of alcohol and drug consumption among young individuals poses significant societal challenges, necessitating effective preventive measures and early intervention strategies. This study suggests a unique method for predicting youth drug use that makes use of deep learning algorithms and eye gaze movement. This work objective is to use eye gaze movement and deep learning techniques to create a predictive model for young people's use of illicit drugs. The study design involves recruiting a diverse sample of young individuals and collecting their eye gaze data through eye-tracking devices. The collected eye gaze data is preprocessed, segmented, and fed into the deep learning models for feature extraction and prediction. Deep learning algorithms, including ResNet50, VGG19 and Alexnet are employed to extract relevant features and predict the likelihood of drug consumption based on the eye gaze data. The findings of this research highlight the significant relationship between eye gaze movement and illicit drug consumption among young individuals. The developed predictive model shows promising results in identifying high-risk individuals and providing a valuable tool for early intervention and prevention efforts. VGG19 outperforms ResNet 50 and AlexNet with highest accuracy of 0.96. It demonstrates the highest overall performance among the three models for the given task.