November 2019
Intermediate to advanced
346 pages
9h 36m
English
We begin by importing the speech recognition library and selecting a list of audio files whose speech we wish to convert into text. Also, we create a list of keywords we would like to automatically detect in these audio files (step 1). The approach taken, of detecting the utterance of these keywords, can be made more robust through stemming or lemmatization, which effectively accounts for variants of the keywords that have the same meaning. For example, Twitter, Twitted, and Tweet would all be detected if this approach is properly implemented. In step 2, we specify that we will use Google's Speech Recognition API to transcribe the audio. Other speech recognition services, such as pocketsphinx, are available as well. We are now ...