A repo designed to convert audio-based "weak" labels to "strong" intraclip labels. Provides a pipeline to compare automated moment-to-moment labels to human labels. Methods range from DSP based foreground-background separation, cross-correlation based template matching, as well as bird presence sound event detection deep learning models!
Other
16
stars
10
forks
source link
Add more detailed Python installation instructions #201
What more detail is required/desired?
We should update the docs to remove python3.8.
If it is not obvious that they need to install the correct python for their OS and architecture, then maybe that person isn't ready to take on PyHa.