Image annotations exported in JSON format use percentages of overall image size, not pixels, to describe the size and location of the bounding boxes.
確定! 回上一頁