Implementation of a mobile supported glasses recommendation system according to human face type using deep learning and internet of things

The rapid functional growth of mobile devices and increasing computational power have created significant opportunities for the application of deep learning algorithms on mobile platforms. Deep learning is an important sub-discipline in the field of artificial intelligence and great progress has been made in recent years. With these advances, deep learning methods have been used to increase data volumes, providing solutions for complex calculations with high levels of abstraction and pattern recognition capabilities. Since these capabilities provide great advantages in many application areas, the coordination of different systems that will work online with mobile devices has become very important. In recent years, the increase in the use of mobile devices and the communication of online systems with each other have become one of the important application areas for the use of deep learning methods on mobile devices. In the thesis study, a photo obtained by using the photo-taking feature of the Android device with the developed mobile software or an image selected from the gallery of the Android device was improved with image processing techniques. The improved image was analyzed using deep learning methods with a different device that works online over the internet, and it was determined which face type the person had within the determined face class types. In addition, the list of glasses types most suitable for the determined face type was displayed to the person with the android software developed within the scope of the thesis. Thus, in addition to recommending glasses to the person with the unique software developed, it saves users both in terms of space and time.