Vosk API is an offline speech recognition API for Android, iOS, Raspberry Pi and servers with Python, Java, C# and Node.
I have a speech to text GUI program using Vosk API that transcripts spoken words to text at the mouse cursors location. It has several features of which I would like to modify and several I would like to implement. Currently, it is using an addon called "Fastpunct" which automatically punctuates the sentences however this causes a huge delay to output the text so I am looking for a different solution. Also, the speech program has a feature to enable "commands" which means whenever you say a form of punctuation such as "Question key" it will translate to "?". Many of the commands are either working periodically or not at all. I would like to speak to someone about possibilities to fix this issue whether it is modifying the existing model or creating a separate one just for the command feature. Furthermore, I have much more future work/ideas that need implemented that we can discuss.
7 freelancers are bidding on average $191 for this job
Hi, thanks for your job posting. Punctuation is rather NLP processing. Vosk API has look ahead function with which I can implement your needs. please contact me. thanks. Anton.