The performance of vision-based systems is severely hampered by foggy weather, particularly when it comes to person detection tasks. The goal of this research is to use a Generic ModelAgnostic Convolutional Neural Network (GMAN) to create a deep learning-based dehazing and person detection system. Using a GMAN dehazing module, the suggested system first eliminates haze from the input image or video frame, improving visibility by restoring lost features. For precise person detection, another GMAN-based model processes the improved frame after dehazing. The dual-task method is appropriate for security systems, driverless vehicles, and surveillance since it guarantees increased accuracy in difficult situations. To prove the model's superiority over current techniques, benchmark datasets will be used for evaluation. In this study, we use the Generic Model-Agnostic Convolutional Neural Network (GMAN) to present an enhanced framework that combines human detection and image dehazing into a single system. The system is made to handle the difficulties of processing images in foggy environments, where sight is impaired and in-depth analysis is necessary. By restoring tiny details lost to haze, the dehazing module improves image quality and clarity. The dual-task technique is appropriate for security systems, driverless vehicles, and surveillance since it guarantees greater accuracy in difficult situations. To prove the model's superiority over current techniques, benchmark datasets will be used for evaluation.