基于STDP的脉冲神经网络模型实验数据集
收藏国家基础学科公共科学数据中心2026-03-28 收录
下载链接：
https://nbsdc.cn/general/dataDetail?id=69bec092bb16e02c49cd82e0&type=1
下载链接
链接失效反馈官方服务：
资源简介：
本数据集包含7,000条测试样本，每条样本为1秒长度的单词语音录音，采样率为16kHz。该数据集具备说话人和环境多样性，显著增强了噪声鲁棒性评估能力，在本课题中用于研究训练和验证基于STDP的脉冲神经网络模型所涉及的基于STDP的语音识别技术，相关性能指标包括准确率AP等。
This dataset comprises 7,000 test samples, with each sample being a 1-second single-word speech recording sampled at 16 kHz. Boasting diverse speakers and environmental conditions, this dataset significantly enhances the capability of noise robustness evaluation. It is utilized in this research to study and validate the STDP-based speech recognition technology for training and verifying spiking neural network models, with relevant performance metrics including accuracy AP and other indicators.
提供机构：
杭州海康威视数字技术股份有限公司
