Driver fatigue is a leading cause of accidents worldwide, resulting in significant loss of life and property and much more. Our study proposes a model for real-time driver state detection, capable of detecting whether a driver is sleepy, drowsy, or alert. Capturing real-time data, our model demonstrates exceptional accuracy in detecting driver fatigue. Notably, the proposed model performs well even in low-light or dim conditions, ensuring its effectiveness in various driving scenarios. Our proposed model has significant suggestion for enhancing road safety, reducing accidents, and saving lives by providing a confident means of driver state detection, contributing to the development of intelligent vehicular movement systems and a reliable driving environment.