hongyang
@ouuan Thanks.
@YuwenXiong I test your code. There indeed exists the two problem. https://github.com/uber-research/UPSNet/blob/2ced987576a98dc16c5a580839b657948644d44d/upsnet/dataset/base_dataset.py#L941 https://github.com/uber-research/UPSNet/blob/2ced987576a98dc16c5a580839b657948644d44d/upsnet/dataset/base_dataset.py#L942 In this place, the small pictures will be padded 255. For the second question, you can check...
@JoyHuYY1412 Hi, I met this problem when using this code. That's because the "crowd instances" are included. You can refer to #33 . Please let me know if you have...
@JoyHuYY1412 I am sorry that I can't remember exactly. But I remember that the performance seems to be worse. The pan_loss's becoming larger is normal. Because when you drop the...
@JoyHuYY1412 Oh, you are right. I was wrong. It's hard to say. It seems that we need the author's help. @YuwenXiong
@zimenglan-sysu-512 I also re-implement the panet using maskrcnn_benchmark（2mlp without ms train， gn， sbn）. But the performance is worse than this repo. If convenient, can you share your code with me?...
@gaussiangit Hi, the semantic results is used in this place. https://github.com/LaoYang1994/PanopticSegmentation/blob/467e3c98e370e24cc9c9ab45be16aa940b1a22cc/src/cal_panoptic.py#L142 It's just a dummy interface just for debug. Actually during the competition, we load semantic model in this place...
@wudongyuan Hi， this file is just the instance-segmentation-json-result. It is generated through your own Instance-seg model (for example, MaskRCNN).
@wudongyuan You can download the trained-model provided by [detectron2](https://github.com/facebookresearch/detectron2) or [mmdetection](https://github.com/open-mmlab/mmdetection) to generate the file you want.
@shwoo93 Thanks for your attention! I have finished the d2-based code, but there are still some bugs. Recently, I'm busy with my graduation and have no time to fix these...