Speech recognition in real-time with punctuation

Question

What is the way to implement speech recognition (voice to text) with auto punctuation? I want to use it to turn a lecture (45 min talk) into text and if possible update the view dynamically.

I tried SpeechRecognizer but it only gives me words without punctuation and stops listening after the first words.

Answer 1

You can use Punctuator, it assigns punctuation to ASR result with the help of a deep neural network.