ARCA23K

ARCA23K是由英国萨里大学视觉、语音和信号处理中心创建的一个音频数据集，包含超过23,000个来自Freesound平台的标签音频片段。该数据集旨在研究开放集标签噪声，通过自动检索和筛选过程创建，以确保研究的可重复性。数据集主要用于音频分类和机器学习研究，特别是在处理标签噪声对模型训练的影响方面。

ARCA23K is an audio dataset created by the Centre for Vision, Speech and Signal Processing at the University of Surrey, UK. It comprises more than 23,000 labeled audio clips obtained from the Freesound platform. Developed to investigate open-set label noise, this dataset was built using an automatic retrieval and filtering workflow to guarantee the reproducibility of associated research. It is primarily utilized for audio classification and machine learning studies, particularly those examining the impact of label noise on model training.

搜集汇总

背景与挑战

背景概述

ARCA23K是由英国萨里大学创建的大型音频数据集，包含超过23,000个来自Freesound平台的标签音频片段，专门用于研究开放集标签噪声及其对音频分类和机器学习模型训练的影响，并通过自动化过程确保研究的可重复性。