Portfolio

Speech to Text application

Type Mobile application
TECHNOLOGIES Java, ReactJS
DURATION 10 months
TEAM 3 developers

The application is designed to convert speech to text in any language for deaf and hard of hearing people. Not only can the app transform speech to text, but it can also save it as a file.

The application is currently being worked on. The next phases of development imply integration with smart glasses and the implementation of a text to speech feature.

Challenges
  1. The Customer wanted to integrate with IBM Watson, however, the system did not support all necessary languages.
  2. The application was designed to convert a voice message into a text. The Customer was going to have a sound file of the voice message as well.
Solutions
  1. The developers performed IBM Watson integration as default. In case IBM Watson does not support the required language, the application applies Google Speech API for conversion.
  2. InputStream А from a microphone was transmitted to object B which read from A and recorded data into inputStream C. At the same time there was a second T which recorded C into .wav-file.
Results
  1. We integrated the developed software with both IBM Watson and Google Speech API. The application enables users to translate the speech into all the languages which are supported by IBM and Google.
  2. Not only can the application transform speech to text, it also records the speech into files.
Global Event Map application
Global Event Map application
Traffic and sales report generation system
Traffic and sales report generation system