![]() ![]() On their Speech Service overview page Google describes how they are able to take audio files of any length, run a machine learning model on it that is trained to turn speech into text and deliver results fast and free (up to 60 minutes) for a set of over 120 languages. The solution: Google Cloud Speech APIĭuring my five minute research I came across the Google Speech to Text API and the connected Google Cloud Services. The audio files from my friend have been >30min and longer and splitting it up into small chunks to get around the limitation was neither ethical nor a fun task. ![]() Most services will offer some free tier solution, where 1 min is free until you get charged (freemium model). This makes all software (I found) out there expensive since a well trained machine learning model has to run on high performing hardware to get the spoken language into text. Its either time intense (manually typing), cost expensive (to hire someone or use a tool) or computing intense (which results into cost expensive). Unfortunately this is an expensive problem to solve. The last sounded like the most sane solution to the problem and so I looked up some existing software that is out there that tackles this issue. hire someone who transcribes the interview for you.listen to the interview and type it off by yourself.It seems like there are 3 ways of getting a spoken interview transcribed: To me this sounded like a solved problem and I was sure that there are some free services out there that could just do this for her but it turned out to be harder than I thought. Run the Script by providing the new location Getting a recorded interview transcribed by the Google Speech API # ![]()
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |