-
Hello,
I found that on the split of the ShanghaiTech dataset, there are a few classes that have data on the test set but not on the train set and vice versa. For example, there's a total of 4 videos …
-
-
Hi everyone :smile:
I'll be running the session on unsupervised learning on Wednesday and I would like to get an idea of who you all are and what your plans are with ML. So if you could put in you…
-
Are there any semi-supervised collections, such as _MemSeg_? If I want to read semi-supervised papers, are there any recommendations? thanks 🙏
-
Dear author, thank you for your excellent work!
I was wandering where do you get the training annotation file of ShanghaiTech Campus dataset? For I see that the original dataset is unsupervised, p…
-
Thank you so much for open-source such solid work. At present, I have done a weakly supervised segmentation task on my dataset only containing the target and background. I learned Teacher-Student Netw…
-
Using MCL_55.15_58.5.pth as the weight, run the max miou:42.71176333180985 obtained by train_mcl.py?
The max miou in the paper is 58.4?
-
Hi there,
We are currently working on a paper in the WSOL domain and plan to compare our results with the following papers authored by your team:
1) Negative Evidence Matters in Interpretable Histol…
-
Thanks for the great job! In this study, both the 2D backbone lseg and openseg necessitate text labels for input. This approach appears more akin to weakly supervised learning (as described in https:/…
-
Hi, I encounter the same problem as in https://github.com/facebookresearch/LAMA/issues/10.
And I found the reason why 2 examples are filtered is that the `obj_label` are `1970s` and `1990s`. And in `…