Supplementary information for the Interspeech 2024 paper:
Liang et al. 2024. Improving Audio Classification with Low-Sampled Microphone Input: An Empirical Study Using Model Self-Distillation.
======
The details of the hyper-parameters we used: Supplementary_information-camera_ready.pdf
The main Python scripts: /scripts/esc_folds.py, /scripts/tau2019_folds.py (to pre-process audio samples), /scripts/kd.py (self-distillation)
Demonstration of the re-sampling outcomes: /librispeech_effects/