海德堡脉冲数据集

海德堡脉冲数据集是由德国海德堡大学物理研究所创建的一个高质量音频数据集，专门用于评估脉冲神经网络。该数据集包含约10,420个英语和德语发音的数字录音，由12位不同年龄和性别的演讲者录制。数据集的创建过程涉及精确的音频对齐和转换为脉冲信号，以适应脉冲神经网络的输入需求。该数据集主要用于语音分类和关键词识别，旨在通过脉冲神经网络提高语音处理的效率和准确性。

The Heidelberg Spiking Dataset is a high-quality audio dataset created by the Institute of Physics, Heidelberg University, Germany, specifically designed for evaluating spiking neural networks (SNNs). It contains approximately 10,420 digital audio recordings of English and German spoken numerals, recorded by 12 speakers with varying ages and genders. The dataset creation process involves precise audio alignment and conversion into spiking signals to meet the input requirements of spiking neural networks. This dataset is primarily used for speech classification and keyword spotting, aiming to improve the efficiency and accuracy of speech processing via spiking neural networks.