Resources from the Gardner Lab
TweetyBERT
TweetyBERT combines a convolutional front-end with a transformer architecture to learn representations of bird vocalizations.
Full source, installation instructions and usage examples are on Github.
Deciphering Canary Song:A Deep Dive into Self-Supervised Learning with TweetyBERT with George Vengrovski
TweetyBERT:Automated parsing of birdsong through self-supervised machine learning with Tim Gardner.
If you use any version of TweetyBERT or its source code, please cite this preprint:
Vengrovski G, Hulsey-Vincent M, Bemrose M, Gardner T. TweetyBERT: Automated parsing of birdsong through self-supervised machine learning. Patterns 2026. https://doi.org/10.1016/j.patter.2025.101491
George Vengrovski, georgev@uoregon.edu
Open an issue on the Github repository.