Computational Audition with Imprecise Labels

This doctoral dissertation presents research on computational audition techniques that can learn from imprecise labels. The work addresses fundamental challenges in audio understanding when training data has weak, noisy, or partial annotations. The thesis covers methods for sound event detection, audio classification, and representation learning that are robust to various forms of label imprecision, enabling practical deployment of audio AI systems where perfectly labeled data is unavailable.

Recommended citation: @phdthesis{shah2024computational, title={Computational Audition with Imprecise Labels}, author={Shah, Ankit Parag}, school={Carnegie Mellon University}, year={2024} }