Your input image dimensions are considerably smaller than what the CNN was trained on and increasing their size introduces too many artifacts ...
確定! 回上一頁