You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
the potentially easier idea could be to try to train a visual network. Not looking at finger positioning on the instrument, though that could be one way (but harder?)
Most of these tutorials have the chord diagrams displayed on the screen.. we could build a system to identify that.
We could generate a training set for this by randomly generating chord diagrams and placing that on top of random frames from videos. Then trying to identify where that chord diagram appears in the video. And once we pick the diagram, we can build a "hand writing classifier" kinda of thing to get the actual chord name.
Could we do audio detection? Use youtube transcript for the video? Use image processing on the video frames?
The text was updated successfully, but these errors were encountered: