Zongwei Li

Sep 10, 2023 | mdpi.com | Zongwei Li |Kai Qiao |Jianing Chen |Zhenyu Li

4. Experience 4.1. Data Set and Experimental EnvironmentIn our research, we utilized a self-assembled dataset to evaluate and train the enhanced model. The dataset comprises a total of 1844 images, encompassing four distinct target types: utilitarian, emotional, normal, and hedonic.