-
Dear Developer,
I'm trying to use the sopa to output the xenium file, so I can open it with xenium.
but I got the error said no cell circles, I actually have this one in my spatialdata (pelase c…
-
Thanks for your work, but we ran into a difficult issue while using it. We look forward to hearing from you. Here is the problem description:
(1) Entering the scorer unet network at 31 epochs, an epo…
-
I am interest in 3D instance segmentation.
I found that you said your code would be release soon.
I wonder when will you release your code?
-
Dear authors, when I tried to use the checkpoint download from ESAM-E under 3D instance segmentation results on ScanNet dataset section in README. I test the result on scannet-mv dataset, but the AP p…
-
-
https://github.com/xtreme1-io/xtreme1
Xtreme1 unlocks deep insights into data annotation, curation and ontology management for tackling machine learning challenges in computer vision and LLM. The p…
-
Dear Alexandre,
I successfully applied ConvPoint to segment point clouds. Specifically, I segmented points into 3 classes (building, vegetation, ground) for urban scenes. Now I would like to perfor…
-
There are different ways for refining existing masks with `micro_sam`.
The easiest option would be to derive point prompts from the centers of the masks and then prompt the model with these points.…
-
Hello! Thank you for your awesome work and your contribution to 3D instance segmentation.
May I ask for your detailed evaluation results on S3DIS?
Did you calculate the AP50 of fl…
-
Dear authors,
thank you for share your great works.
When I read the code, I found this code use the Mix3D data augmentation, this augmentation achieves good performance on MinkowskiNet back…