These approaches [61,. 62, 38, 39] address NR-IQA as a learning-to-rank prob- lem, where the relative ranking information is used during the training. Zhang et ...
確定! 回上一頁