Ellaft/pc-fault-real-dataset

该数据集是一个多模态数据集，包含音频和图像特征，用于故障检测任务。音频特征采样率为16000Hz，图像特征为图像数据。故障类别包括正常操作、启动失败、风扇过热、存储故障和系统崩溃5种类型，对应标签为0到4。数据集包含训练集（1750个样本）、验证集（375个样本）和测试集（375个样本），总大小约1.86GB。数据来源可能涉及设备故障监控场景。This dataset is a multimodal dataset containing audio and image features for fault detection tasks. Audio features have a sampling rate of 16000Hz, and image features are image data. Fault classes include normal operation, boot failure, overheating fan, storage failure, and system crash, corresponding to labels 0 to 4. The dataset includes a training set (1750 samples), validation set (375 samples), and test set (375 samples), with a total size of approximately 1.86GB. The data source likely involves device fault monitoring scenarios.