Speech recognition and transcription software
The use of speech recognition and transcription software at the courts could facilitate compliance with the deadlines relating to the obligation of putting decisions and minutes into writing, and would also result in more efficient use of the working time by reducing the time required for transcription.
Speech transcription programs instantly transform the speech of the person speaking into editable text. The programs are capable of transcribing the text simultaneously with the speech or from audio recordings, and can also recognise acronyms, abbreviations, symbols, Arabic and Roman numerals, fraction numerals and references to legislation. The text transcribed by these software can be imported to a text editor program.
Starting in 2016, several courts tested two speech recognition and transcription programs – Globalspeech and Speechtex – simultaneously, and the testing period ended at the end of 2017. During the testing, the pilot courts continuously compared the ease of operation of the software, the time required for learning them, their rate of correct recognition of speech and the time saved.
Starting from 1 June 2017, the President of the NOJ mandated the President of the Zalaegerszeg Regional Court to test, procure and communicate speech recognition and transcription programs at the courts. During the testing, the users at the Zalaegerszeg Regional Court acquire new experiences and continuously compare the ease of operation of the software, the time required for learning them, their rate of correct recognition of speech and the time saved. The users forward
the development and advancement possibilities and practical solutions to the developers. The test results were presented at the courts in the scope of a roadshow, after which the courts indicated whether they intended to procure speech recognition and transcription programs.
In 2018, 726 speech recognition and transcription software were purchased from the courts’ own budgetary sources.