Q: How do you pick your Spectogram parameters and window length?
Answered in stream.
Q: What software do you use for labelling audio data?
Answered in stream.
Q: Could we just compare detected BPMs to actual BPMs for model evaluation?
Answered in stream.
Dusan Smitran 👏 Maciej Słomka 👏 Jean-christophe loeb 👏 Juho Vikman 👏
Akshay Zolekar 👏 Nikos Michas 👏 Christoph Gietl Jon Nordby: The link "Audio Classification with Machine Learning (EuroPython 2019)" in your slides seems to be broken ("Video unavailable"). Jon Nordby: Apparently a typo in the URL
I want to measure my Keystrokes Per Minute :D
Dusan Smitran: I think measuring Keystrokes per Minute could be fun. A nice thing is that one could use a keystroke counter on the machine to provide labeled data automatically. Not sure if an acoustic solution would compete with one that is tied into the OS input facilities though, but it is an interesting challenge!
Christoph Gietl: ah thanks, that is a good point. It was a local video. Will try to see if I can link to Youtube versions
Hi everyone. Thanks to all that attended the talk Sound Event Detection with Machine Learning! If you have any questions I am around for a bit to discuss. We can do this for example in the Breakout 4: Parrot room :)
please share the link to the code shown in the presentation
The project I talked about is found here, https://github.com/jonnor/brewing-audio-event-detection
GitHub - jonnor/brewing-audio-event-detection: Tracking beer/wine using Audio Event Detection with Machine Learning
- GitHub Tracking beer/wine using Audio Event Detection with Machine Learning - GitHub - jonnor/brewing-audio-event-detection: Tracking beer/wine using Audio Event Detection with Machine Learning