Thinking about the music analysis strategy:
- One hosting platform at a time, in batches, starting with Soundcloud
- Only feeds that use the music category tag
- Exclude Anchor because they don't allow music anyway
- Only analyze enclosures with durations less than 20 minutes because songs are hardly ever that long
@Lehmancreations This is what I’m tinkering with now: https://github.com/fmqa/samplecnn-speech-detection …but there are quite a few on GitHub. Mostly Python, which stinks for me, but I’ll adapt I guess.
Intended for all stake holders of podcasting who are interested in improving the eco system