New BECCA video: Hierarchical spatio-temporal feature learning in audio data

[Brandon Rohrer]

Hi all,

It's been a while in the making, but here is a 60 second presentation of some research results:

https://www.youtube.com/watch?v=teXnVbjjvCU&list=PLF861CC4C40439EEB

BECCA listens to The Hobbit and creates several levels of spatio-temporal features (actually tonal-temporal features, since it's audio data). The code is untagged, but you can get it from the latest commits at:

https://github.com/brohrer/becca
https://github.com/brohrer/becca_world_listen
https://github.com/brohrer/becca_tools_control_panel

I didn't commit the raw data files, but I can get you copies if you want. Just let me know.

I'm planning to release a few more results videos in coming weeks. I'll share links to each with the group.

Cheers,

Brandon

[Matt Chapman]

Cool! Thanks for sharing!

All the Best,
Matt Chapman

--
You received this message because you are subscribed to the Google Groups "BECCA_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to becca_users...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.