ZDY

Ok, thank you!

Where are the query images from during the training phase? Are the query images from coco val2017 too in the training phase?

In your work, you visualized the bounding boxes with scores larger than 0.8. What are the scores related to? Is it related to both IOU and categories?

I am a postgraduate of China University of Geosciences. I have been reading your paper and code rencently.I really appreciate your work.I have download your new dataset,but I cann't apply...

Is the generated pseudo_mask in result/ir_label folder?

> > Thanks for your work. How to replace the ground_truth annotation in PASCAL VOC 2012 with the generated pseudo_mask? > > Have you solved the problem ? 替换文件夹即可，将原先存放真实标签的文件夹用来存放你生成的伪标签。

> > > > 请问通过run_sample.py生成的图像的伪标签(Pseudo-Mask)是result文件夹下的ir_label吗？ 不是，run_sample.py生成的伪标签存放在sem_seg文件夹中，ir_label存放的文件用于是训练IRNet的标签，具体细节可参考IRNet论文和代码。


关闭错误之后，预览的图片也不能放大和缩小

I have addressed this problem. Updating to the latest version of Element Plus can resolve issues because it includes bug fixes