How to reuse existing annotated image quality datasets to enlarge available training data with new distortion types.评价结果

评估详情