Xingyi Yang

By the way, it would be most appreciated if you may introduce this work to your friends and help us on promoting this project. Best.

Dear @BloomBerry , We are communicating with the `mmdet` team on the version update support. Stay tuned. Best.

Dear @yuan738, Thank you for your insightful question. Currently, the semi-supervised method is heavily dependent on a large batch size, and as a result, reducing the number of GPUs could...

Dear @zimenglan-sysu-512 , I recommend increasing the ratio of labeled to unlabeled samples, such as a `1:1` ratio. Currently, the learning rate is based on `8 labeled samples (1 labeled...

Dear @zimenglan-sysu-512, Yes, setting data.sampler.train.sample_ratio to `[2, 4]` or `[3, 3]` should work well. The first number represents the number of labeled samples, while the second number represents the number...

We only has a file called `configs/consistent-teacher/consistent_teacher_r50_fpn_coco_720k_fulldata.py`. No config provided for 360k training for full data.

Dear @zimenglan-sysu-512, Apologies for any confusion caused. I wanted to inform you that the file `configs/consistent-teacher/base.py` has been renamed to `configs/consistent-teacher/consistent_teacher_r50_fpn_coco_180k_10p.py`. I have also updated this change in the README....

Dear @zimenglan-sysu-512, 1. In our experiment, we decided not to decay the learning rate. Surprisingly, we observed that using a fixed learning rate resulted in higher performance compared to using...

Dear @zimenglan-sysu-512 The results you provided are amazing. It would be very helpful if you could share the config and checkpoint. This experiment could be extremely useful for people with...

Dear @Re-dot-art and @Code-of-Liujie, Could you please provide details about the specific issues you're encountering with the environment configuration? Our code was originally developed during the mmcv1.0 era. Since then,...