In today's digital era, the importance of human–computer interaction (HCI) is growing for various applications such as entertainment and medicine. Natural and more comprehensible interfaces such as voice commands, gestures, and touchscreens, which are more easier to understand, have fully replaced traditional methods of input such as keyboards and mouse. This paper outlines a thorough method for detecting gestures in low-light conditions using advanced image processing and deep learning models. The proposed approach includes deep learning algorithms to enhance image quality, segment the hand gestures, extract features, and accurately classify hand gestures in challenging low-light conditions. Input images are enhanced with the MIRNet model, which effectively retrieves information and increases visibility in low-light conditions. BASNet is used for segmentation of the image. Using MobileNet algorithm, gesture classification is made possible by extracting feature from segmented images. These components operate together in a cohesive structure to provide strong performance in all kinds of situations. Experimental tests have shown that the suggested method could effectively identify hand gestures during a low-light environment. The functionality of the system in dynamic circumstances is illustrated via the inclusion of media player control using JavaScript user-defined functions and Flask microweb framework in Python, demonstrating its real-world application. The overall goal of this project is to provide a complete solution for low-light gesture identification by using cutting-edge approaches.