Research
Notes from the speech lab
- 4 min read
Speech models learn physics
We examined convolution filters across six speech systems and found that training repeatedly produces Gabor-like shapes predicted by signal theory.
Read → - 3 min read
What speech encoders hear
We mapped every layer of four speech encoders and found that emotion, speaker identity, and sentence content settle at different depths.
Read → - 1 min read
Hello, world
Why oruk is starting a public speech research journal, what we plan to publish, and the standard we want the work to meet.
Read →