search

fschmid56 / EfficientAT

This repository aims at providing efficient CNNs for Audio Tagging. We provide AudioSet pre-trained models ready for downstream training and extraction of audio embeddings.
MIT License
215 stars 40 forks source link

AudioSet temporally-strong labels #30

Open knma1992 opened 3 months ago

knma1992 commented 3 months ago

Firstly I would like to state that this repo is great, so many models, all in pytorch and getting them to work on my machine was very easy.

Have you tried fine-tuning the models on the temporally-strong labeled subset of the AudioSet dataset?

fschmid56 commented 3 months ago

Hi, thanks for your interest and kind words!

Indeed we are currently working on this. I'm experimenting with a frame-wise version of DyMN and fine-tuning it on AudioSet Strong. I will update this repository with new models as soon as the experiments are conclusive.