html PERCEPTION AND AI: FACIAL RECOGNITION FOR VISUAL DISABILITY – Perception: Creator of visual assistants for the blind

PERCEPTION AND AI: FACIAL RECOGNITION IN THE SERVICE OF VISUAL DISABILITY

PERCEPTION AND AI: FACIAL RECOGNITION IN THE SERVICE OF VISUAL DISABILITY

At the forefront of future technologies, the Perception team has combined artificial intelligence with an embedded system. This technological solution called Nélo is now used in the industrial sector. Today, it is available to help the blind and visually impaired to move around, read, recognize objects and faces with a facial recognition system. Discover in this article the face and object recognition capabilities of Nélo, the intelligent personal assistant created by Perception!

 

Facial recognition, a technology adapted to the blind and visually impaired

The technology of facial recognition is based on theartificial intelligence. It allows a person to be identified and authenticated. Facial recognition uses, in addition to'Artificial intelligence, biometrics, 3D mapping and Deep Learning which analyzes faces to be able to compare them. There facial recognition is used in the fields of security, marketing and health. In the security and commerce sectors, facial recognition sparks numerous debates on the protection of freedoms and privacy. However, in the health sector, it represents new opportunities for diagnosing diseases and supporting patients.

It is with the ambition to assist people suffering from eye diseases, blindness and visual impairment that the team of engineers at Perception in Marseille, has developed a intelligent visual and audio personal assistant which uses a solution of facial recognition. Nélo, better than connected glasses will revolutionize the lives of the visually impaired.

 

The intelligent assistant Nelo automatically recognizes faces

L'Nélo portable personal assistant uses Deep Learning to automatically recognize faces and objects. The algorithm developed by Perception is learning and adaptive, that is to say that it is trained to recognize faces from videos captured by a small high-definition camera. At most Nelo is used, the more its learning model improves to become smarter day by day. Also, a pocket computer with great computing power analyzes the images collected by the camera in real time.

The person suffering from blindness or visual impairment receives, using an extra-aural headset, a description of the faces or an alert if the person encountered is already registered in the database. Only the device user hears the audio description. Audio notifications do not resonate with usual means of hearing. So noises and all sounds continue to be heard.

 

Facial recognition improves social life and safety for blind and visually impaired people

Recognizing family members, friends, work colleagues, is a natural thing for an ordinary person. Indicating with a smile or a short phrase that you are happy to meet them improves social relations. However, visual disability makes social relationships complex.

Indeed, non-verbal language, such as facial expressions, is not perceptible to the visually impaired. Nelo, promotes communication with others. Through audio descriptions of faces and expressions, a mental image is formed.

The visually impaired individual therefore creates a personal representation of the face of the person they meet.

In addition, the AI ​​developed by Perception identifies and describes facial expressions. Nelo then becomes a fantastic tool to support socialization. Likewise, this assistant is a tool to help with self-construction for children and adolescents. By making it possible to identify and authenticate the faces of loved ones instantly in any moving environment (a crowd, a group), the facial recognition reinforces the feeling of security of blind and visually impaired people. Nélo is undoubtedly the ideal intelligent solution for blind and visually impaired people. Nélo is an autonomous visual assistant or guide for the blind, without internet and more powerful than connected glasses ( Panda, E-sight…)