A Real-Time System to Assist Blind People Through Face Recognition and Emotion Detection

As per the recent report by the world health organization (WHO), there are 2.2 billion people with various kinds of vision impairment. Around 50% of people with visual impairment are entirely blind and require the support of technology to make them more accessible in their life. In this paper, we propose an assistive system to help blind people. In general, blind people will understand their friends and relatives through their voice, smell, or rhythm of their footsteps. But the blind may not be able to judge emotions too quickly since facial expressions demonstrate the emotions of individuals. This work proposes a system that will use a combination of face recognition and an emotion detection approach to recognize the persons and identify the emotions for assisting the blind through personalized audio outputs. The system will identify the person and their emotion in those who are approaching the blind, and it will give an audio description about them well in advance to the blind so that he/she can fine-tune the interactions in a better way. Certainly, in some situations, the blind may want to interact with strangers, and in that situation, our system will help the blind by giving an audio description that includes approximate age, gender, other physical parameters, etc. A detailed experimental study is carried out with randomly selected video samples and the results show that the system works well in real-time.