Scientist to Work With Philips on Kazakh Speech to Text AI

Kazak ASR Telegram chatbot is already in service

By Marina Rakhmatulloyeva

Qazaqstan Monitor: Scientist to Work With Philips on Kazakh Speech to Text AI

Dutch tech company Philips offers Kazakh scientist to collaborate on Kazakh-language speech to text AI technology.

Orken Mamyrbayev has developed a speech recognition system in the Kazakh language that converts voice messages into text. Mamyrbayev is a researcher at the Institute of Information and Computing Technologies of the Committee of Science of the Ministry of Education and Science of the Republic of Kazakhstan.

AI, linguistics and the Kazakh language

The intelligent system "Kazak ASR" is already working in the social network Telegram as a chatbot. It is planned that the development will also be integrated into products of the famous Dutch company Philips and will soon appear on the market. The project was developed thanks to grant funding from the Ministry of Education and Science.

Mamyrbaev devoted his scientific activity to the development of the state language and is now engaged in projects related to artificial intelligence and computational linguistics.

"Kazak ASR" will not only recognize voice signals but also convert them into coherent text. The developed system is already working in Telegram as a chatbot. To test the system, you need to press and hold the voice message button. Once the recording is complete, the bot will turn your voice signals into text. My team and I have developed a voice recorder, which can be turned on during a meeting or a meeting, and after its completion, the device will prepare minutes in the Kazakh language."

At the moment, the act of introducing the implemented technology of Kazakh speech recognition in Philips products is being prepared. As part of the collaboration, the Smart Mike Duo microphone from Philips with speech recognition, which automatically converts speech into a printed text, will be presented to the Kazakh market. The device is capable of recognizing two different voices that sound simultaneously and automatically converts speech to text on two different channels.

"Philips has offered to collaborate and further promote the product. If we talk about the algorithm step by step, the software is loaded into the computer. Then, we connect a microphone and start talking. The device automatically recognizes two different voices and outputs text in the Kazakh language in two blocks. Such a microphone is useful during working meetings for protocols, where you may need to save information in the form of text."

Only 30% of an article published on The Qazaqstan Monitor website may be used with a mandatory hyperlink provided to indicate the original source. To re-publish the full article, written permission from the editorial is required.