When I use the coco dataset (133 classes) to make predictions on my images, I find that it merges some of the targets. For example, for three houses, it outputs building-other-merged. while this is a correct answer, I would like it to output one category for each identified target, i.e., for three houses, it outputs building three times. another example, the model will merge pavement and road as pavement-other-merged But I want it to output road and pavement. what should I do? Thanks in advance!
When I use the coco dataset (133 classes) to make predictions on my images, I find that it merges some of the targets. For example, for three houses, it outputs building-other-merged. while this is a correct answer, I would like it to output one category for each identified target, i.e., for three houses, it outputs building three times. another example, the model will merge pavement and road as pavement-other-merged But I want it to output road and pavement. what should I do? Thanks in advance!