Multi-axis quality evaluation of data augmentation for perovskite solar cell datasets

A three-axis framework evaluates augmented data through predictive response, data compatibility, and structural preservation, providing a systematic approach for assessing data augmentation quality and reliability.