search

Wu0409 / DuPL

[CVPR'24] DuPL: Dual Student with Trustworthy Progressive Learning for Robust Weakly Supervised Semantic Segmentation.
63 stars 3 forks source link

Regarding the issue of reproducing results #6

Open InMyDreammer opened 2 hours ago

InMyDreammer commented 2 hours ago

Dear Author, train.log I have followed your experimental setup and trained the model, but the final results are significantly different from those reported in your paper and log files. Below is my training log file for your reference.

Wu0409 commented 2 hours ago

Hey, thank you for your interest in our project! 😊

I have checked the training log, and it seems that the experiment setting should be correct.

What I can confirm is that the initialization of the dual student does have an impact on final segmentation results, so randomness may cause some performance gaps, but not as much as 2% 😣.

If possible, can you modify the random seed to see if there is a change in the performance gap?