The viewer is disabled because this dataset repo requires arbitrary Python code execution. Please consider
removing the
loading script
and relying on
automated data support
(you can use
convert_to_parquet
from the datasets library). If this is not possible, please
open a discussion
for direct help.
Part of MONSTER: https://arxiv.org/abs/2502.15122.
AudioMNIST consists of audio recordings of 60 different speakers saying the digits 0 to 9, with 50 recordings per digit per speaker [1, 2]. The processed dataset contains 30,000 (univariate) time series, each of length 47,998 (approximately 1 second of data sampled at 44khz), with ten classes representing the digits 0 to 9. This version of the dataset has been split into cross-validation folds based on speaker (i.e., such that recordings for a given speaker do not appear in both the training and validation sets). AudioMNIST-DS is a variant of the same dataset downsampled to a length of 4,000.
[1] Sören Becker, Johanna Vielhaben, Marcel Ackermann, Klaus-Robert Müller, Sebastian Lapuschkin, and Wojciech Samek. (2024). AudioMNIST: Exploring explainable artificial intelligence for audio analysis on a simple benchmark. Journal of the Franklin Institute, 361(1):418–428.
[2] Sören Becker, Johanna Vielhaben, Marcel Ackermann, Klaus-Robert Müller, Sebastian Lapuschkin, and Wojciech Samek. (2024). AudioMNIST. https://github.com/soerenab/AudioMNIST. MIT License.
- Downloads last month
- 71