Fine-Tune the Pretrained ATST Model for Sound Event Detection
https://arxiv.org/abs/2309.08153