One day Saverio and Luca, ordinary Italian students, were returning home from university and met a blind man. He couldn't get to the bus stop and asked for help. The young people did not remain indifferent and responded to the request.
A conversation ensued, and it turned out that this is only a small part of the problems that the visually impaired have to live with. So, the blind man told about how he draws a map of the area in his imagination and moves around the corners of the building. However, he is forced to stop in front of traffic lights, because many of them do not emit signals.
That day, Saverio and Luca realized that their specialization in robotics and computer vision could improve the lives of people whose eyes cannot see. That's how the Horus wearable assistant project was born. The device observes, recognizes and describes a person's environment.
Horus consists of two parts: a pocket data center and a headset that is fixed at eye level. A meter-long cable connects them together. The mini-PC is built on the basis of the powerful NVIDIA Tegra computing platform, a high-capacity battery is responsible for its power supply. The headset includes a dual camera, microphone and speaker.
The user can activate one of the functions of the gadget using several buttons located both on the headset and on the pocket unit. The buttons are easy to distinguish due to the individual shape. In some situations, Horus takes the initiative and triggers an automatic action if he believes that he can benefit right now.
Horus Functionality:
Reading the text. The device recognizes printed texts, including on uneven surfaces. The software quickly captures the entire page and sends a signal about it — now you can lower the book and turn the camera in a different direction.
Face recognition The gadget distinguishes people from the situation and makes it clear whether this person is familiar or not. Adding new faces to the database takes seconds.
Object definition. Thanks to three-dimensional perception, Horus distinguishes objects that are similar in size and shape, such as a milk carton and a juice box. The details of the object are described by the voice assistant.
Specifying the route. When moving, the stereo camera promptly warns the user about an obstacle on the way: the closer it is, the stronger the signal will be.
Description of the environment. Due to modern achievements in the field of artificial intelligence, Horus talks about what gets into the camera lens. For example, the device can describe a real landscape or a photo.
It should be noted that Horus transmits voice prompts through bone conduction. Thus, the sound does not enter the environment and cannot be heard from the outside. At the same time, the developers have worked hard on the ability to connect personal hearing devices.
Today, the gadget speaks English, Japanese and Italian. In the future, other European languages will appear. The device is being tested by a limited number of volunteers from Italy. Surely there is a man among them who could not find a bus stop without help before. Now he can do it.
