Image recognition model integration in Unity and CNN Tester app for image recognition

This study proposes a novel method for incorporating deep learning models into the Unity game engine for binary image classification, which led to the creation of the user-friendly mobile app “CNN Tester.” The research trains, validates, and assesses the chosen deep-learning models using a large dataset of dog and cat photos. The trained models can be seamlessly integrated into an interactive application thanks to the cross-platform deployment and real-time rendering capabilities of the Unity game engine. With its simple user interface, the “CNN Tester” program enables people to predict image labels by scanning images with their mobile devices. The app effectively handles predictions, improving accessibility and engagement for users. The study’s findings show how well each integrated model performs, is accurate, and is efficient when used with “CNN Tester” and Unity. The research offers insights into their advantages and disadvantages, helping to promote user-friendly image recognition systems and the seamless integration of deep learning models with practical applications. Researchers, programmers, and educators interested in the possibilities of deep learning and Unity integration will find the proposed methodology useful.