metadata
language:
- da
tags:
- Sorting
- Classification
- Trash
size_categories:
- 10K<n<100K
The dataset can be used to train a model for trash sorting, which consists of five categories of labels/trash audio samples: Metal, Glass, Cardboard, Plastic and Noise. The model created from this data set uses MFC and classifier - with this the accuracy of the model is 90,26%.