Altyngul Kubisenova, an assistant at the Department of Industrial Management and Energy Economics of Novosibirsk State Technical University, together with the team developed the application "Abtseyka", which will help the speech development of children as well as people who have suffered a stroke. The peculiarity of this development is the use of a neural network that gives feedback to the child.
"Our development team has a social educator who has created his own interactive books for developing a child's speech. Based on these results, we decided to develop a specialized application that develops speech through a game. It is based on a neural network that can diagnose a child’s speech, choose a program of games for development, and allow parents and speech therapists to track progress. Mobile games in the app recognize speech and motivate the child to pronounce sounds independently. The exercises created on the basis of modern techniques develop cognitive-speech functions and help to correct defects," Altyngul Kubisenova said.
To work in the application, parents are registered, then through the games, the child’s speech is diagnosed, which allows us to establish the level of development of the speech apparatus. Based on the data obtained, artificial intelligence makes an individual development program in the form of a set of specific games. Parents can track the stages of learning, and the progress of the child. If there are no obvious improvements, the application contains a database of speech therapists to whom you can ask for help.
"The game proceeds as follows: the voice assistant displays the image elements on the screen, the child must assemble a complete picture, and then the voice assistant asks the child to voice what happened. If the child mispronounces a word incorrectly or indistinctly, the voice assistant repeats it until the child says it correctly. Thus, the student hears and attempts to repeat the correct pronunciation. If the child only pronounces words but does not speak yet, then he is asked to pronounce not a word but certain sounds," Altyngul Kubisenova added.
At the moment, the application is designed for children from 3 to 12 years old — the age when speech is launched. The next step of the developers is the development of an application for use by people who have suffered a stroke and other diseases that lead to a malfunction of the speech apparatus.