Computer Engineering Students Developed System for Visually Impaired
Computer Enginering 4th year students, Erin Avllazagaj, Hamza Khan ve Gündüz Hüseynli developed a system that will help visually impaired people, as part of their senior design project in Bilkent University. The project is supervised by Ibrahim Korpeoglu, faculty member in Computer Engineering Department. The system includes an Android mobile app, a device that is wearable around neck, and a backend server running in the cloud. The system allows visually impaired people to take pictures via voice commands, and the system analyzes these pictures and detects and recognizes the faces using AI and machine learning techniques. The recognized faces are read back to the user via the mobile phone or the devices worn around the neck. To do this, the system is trained using a database of labeled pictures. If the database of the user is not enough to recognize, the databases of his/her friends can also be used. Hence the system has a social media feature as well. The classifiers of the user and his friends can collectively be used to do more accurate and faster recognition. The system can also read the words seen in pictures. Additionally, the system can find and read back to the user 5 keywords that are associated with a given picture. The system can also answer some simple questions regarding a picture, like number of people in the picture, etc.
Newspaper articles (links):