Personal Accomplishments
Worked on the proposal slides at the start of the week. I worked on the Genre classification part of the project. This was mainly looking into the K nearest neighbors algorithm.
The algorithm currently gives me a 70-73% accuracy rate on the test dataset which may or may not be sufficient. I looked into the CNN algorithm, and it gives a much better accuracy rate of around 92%. I will look into this over the course of next week.
- https://colab.research.google.com/drive/1hctVbgbCxK8SNuVWfW2e4kA7DtC2hZNC (This is the file where you can see it run on GTZAN dataset)
On Track?
I was sick for most of the week, so was not able to work on classes as much. According to our Gantt Chart, I think I am still on schedule because we are not necessarily training using Spotify features. This might be a step for the future if we see the audio processing metrics are not enough to generate a holistic picture of the music, and we might need spotify to extract musical traits that we are not able to.
Goals for Next Week
The deliverables I will hope to complete by next week are to look into the Signal processing aspect a little more, and run a similar version of CNN too, and see whether the accuracy is truly better as claimed. The Signal processing would include settling on a few characteristics that I can feasibly extract, and see how the python library interfaces with the music files a little more. In the colab file, I used librosa to just stream a snippet of the audio, but here we want to be able to stream it as real time as possible.