zju3dv / gcasp Goto Github PK

Hi, thanks for releasing the great work! However, there is a problem when I generate the primitive annotation. When the program runs to 'mug', it will be interrupted as shown below. I don't know what caused this. Can you provide the primitive annotation you generated?

加载了预训练模型进行测试，跟文章中的结果有差距，请问是哪里复现错了吗？

I'm very impressed with your work! Can you provide some rough results about time consuming?
If possible, pose estimation part and mesh reconstruction part separately. Thanks a lot.

I used the pre-trained weights and pre-trained semantic primitive weights provided by the source code for verification. The results on the Real dataset are basically consistent with the original paper, but the IoU50 and IoU75 indicators on the CAMERA dataset did not match the requirements mentioned in the text. What is the reason for this?

Hi, I'm wondering if the mesh files are contained in the result generate by eval.py, or I need to find ways to generate them manually?
Also, I met a problem when I try to generate result by instruction provided in README, which the eval.py only generate empty folder, could you kindly tell me where is the problem?
Thank you!

Hi, thank you for sharing, but i have the question about how the data_lists/ is generated？

Thank you for sharing，I want to replace the 3dgcn module, is this idea feasible？

Hey I was testing the reconstruction quality of the byproduct reconstruction. The meshes of the other objects look good to me except for the Can class, which always gives out tiny shapes.

I think the pose and scale are correctly applied to transform mesh from canonical space to camera space, as the other objects look very reasonable. I suspect there are some problems with the pre-trained DualSDF-weight, as I observe that during shape optimization in the optimize_shape_ransac function, the decoded shape from the feature is always tiny.

could you kindly see what is going on with the mesh reconstruction?

Here are some examples:

Thank you!

Hi, Thanks for this nice work.
Could you provide pre-trained weights?
Thank you!