I like this a lot. I've been fascinated by Suzuki Toshitakas work on mapping bird sounds to syntax and want to experiment in that space too.
Ive been assisting at a wild bird rehab but not until I got pet birds (released pet birds that no owner could be found for) did I realize they make these extremely faint sounds to each other that I can sometimes just barely make out when I'm right next to them but are not the other quiet humming they make.
My mic can capture those sounds sometimes, but I don't know how to analyze for example 24h of recording in the cage to find slight variations to background noise. It doesn't have to be real-time and not bird specific (want to capture sounds they make that doesn't register as bird in the models).
If anyone has a suggestion please point me in any direction you know of. Audio is pretty new for me.
If you dig this idea, highly recommend the Cornell Merlin Bird ID app (likely built on the exact tech used here). Use it on my walks every week and it spots some wild stuff (it even correctly identifies birds who are being mocked by a Mockingbird). Helped me learn the Eastern Blue Bird call (personal favorite) so I know when they're nearby.
I wanted to do something similar to this, then I started doing some research on birds in general, and those in my locality, then I started learning about Audio and spectograms and Nyquist Theorem and many other interesting audio stuff.
Then I started going through the Intro to Conservation Bioacoustics by Cornell course, and started watching Bioacoustic Talks by the K. Lisa Yang Center cornell center.
And now I am almost at the point where I cant start manually tagging audio sets, for target species so that I can train custom classifiers to identify birds in Rwanda which are poorly detected by birdnet.
TLDR: Being jobless can lead you into interesting ventures.
Thanks for sharing these resources and your story! I followed a very similar path, and ended up doing a biodiversity related MSc, with my dissertation being a custom classifier for poorly detected species in Príncipe. BirdNET and Perch are phenomenal achievements, but struggle in regions where, ironically, most of the world’s biodiversity is. What you’re doing for Rwandan species is so important!!
Excellent kachō-e prompt - working on something similar and found it hard to get the right balance between sharp outlines and watercolors, and especially plant morphology is dicey (eg plants like Cacao that fruit from the trunk instead of branch tips).
Did anyone come across projects that also nail that aspect well?
I like this a lot. I've been fascinated by Suzuki Toshitakas work on mapping bird sounds to syntax and want to experiment in that space too.
Ive been assisting at a wild bird rehab but not until I got pet birds (released pet birds that no owner could be found for) did I realize they make these extremely faint sounds to each other that I can sometimes just barely make out when I'm right next to them but are not the other quiet humming they make.
My mic can capture those sounds sometimes, but I don't know how to analyze for example 24h of recording in the cage to find slight variations to background noise. It doesn't have to be real-time and not bird specific (want to capture sounds they make that doesn't register as bird in the models).
If anyone has a suggestion please point me in any direction you know of. Audio is pretty new for me.
If you dig this idea, highly recommend the Cornell Merlin Bird ID app (likely built on the exact tech used here). Use it on my walks every week and it spots some wild stuff (it even correctly identifies birds who are being mocked by a Mockingbird). Helped me learn the Eastern Blue Bird call (personal favorite) so I know when they're nearby.
I wanted to do something similar to this, then I started doing some research on birds in general, and those in my locality, then I started learning about Audio and spectograms and Nyquist Theorem and many other interesting audio stuff.
Then I started going through the Intro to Conservation Bioacoustics by Cornell course, and started watching Bioacoustic Talks by the K. Lisa Yang Center cornell center.
And now I am almost at the point where I cant start manually tagging audio sets, for target species so that I can train custom classifiers to identify birds in Rwanda which are poorly detected by birdnet.
TLDR: Being jobless can lead you into interesting ventures.
* Nyquist Theorem. https://www.youtube.com/watch?v=IZJQXlbm2dU
* Intro to Conservation Bioacoustics https://www.birds.cornell.edu/ccb/pam-materials
* Bioacoustic Talks https://www.youtube.com/@CornellSounds
Thanks for sharing these resources and your story! I followed a very similar path, and ended up doing a biodiversity related MSc, with my dissertation being a custom classifier for poorly detected species in Príncipe. BirdNET and Perch are phenomenal achievements, but struggle in regions where, ironically, most of the world’s biodiversity is. What you’re doing for Rwandan species is so important!!
Do you think the same could be used for other cases? I'm thinking about detecting problems with cars (vehicles) just by the noise they make
How far has your thinking taken you? This has piqued my interest on how many fields audio can be used to solve problems
Excellent kachō-e prompt - working on something similar and found it hard to get the right balance between sharp outlines and watercolors, and especially plant morphology is dicey (eg plants like Cacao that fruit from the trunk instead of branch tips).
Did anyone come across projects that also nail that aspect well?
Upvoted for the Theobroma cacao reference. Made me curious about what you are up to.
I'm enjoying tracking the local wildlife with my bird listening station.
There's also an excellent alternative to BirdNet-Pi that runs well on non-Raspberry-Pi machines: https://github.com/tphakala/birdnet-go
Birdnet-go is really good and actively maintained. Shout out to tphakala.
Wholesome