Hi all - I'm mentoring a student who is blind on a bioacoustics project to classify killer whale calls. Anyone out there have any experience with or ideas about making spectrograms and/or acoustic measurements accessible to unsighted people? They are doing awesome classifying by listening - just need to think about next quatitative steps. The student is excellent at advocating for themselves and solving problems but would love extra input from the hivemind
6 April 2023 12:26pm
I've got some major vision problems myself so I can relate to their problem. I want to do call recording for cryptic nocturnal birds mostly, but analysing the call recordings has proved difficult.
.I have looked at the Arbimon site, but the process is rather complex to upload files and create call templates. Especially when I can't navigate computer screens as well as I used to. I recently found out about the BirdNet site from Cornell Uni. This uses AI to identify calls against a library of known species. I've processed some recordings and the process is ridiculously easy. Unfortunately. The IDs are mostly US based birds. There doesn't seem to be a way to add your own known calls to identify in recordings, but they are open to collaboration to help develop the system. They even have the app software available for download from GitHub. It could be worthwhile to see if you can modify their software for your purposes.