Novosibirsk: Students from Novosibirsk State Technical University have developed an artificial intelligence-powered video navigation application designed to help blind and visually impaired individuals move around cities more safely. The system requires only an Android smartphone and a chest-mounted camera.
According to Emirates News Agency, the application uses a neural network to process real-time video, identifying obstacles and providing audio alerts. It is powered by a convolutional neural network trained on labelled datasets and is capable of detecting objects up to five metres away. The system responds swiftly to sudden hazards such as vehicles or closing gates.
The app processes visual data in just 300 milliseconds, helping prevent collisions. It works without an internet connection, requires no subscription, and is compatible with entry-level smartphones.
Planned updates will enable voice commands, object identification by name, and user-led AI training to recognise new items. The project recently
won first place in a national student IT competition, with final testing set to take place ahead of its public release.
