Closed mahsa1363 closed 4 years ago
Thanks for using and sorry for the late reply! This is a file path issue and we have fixed it in the latest commit.
thanks for your reply. in folder Data there is merge_ambi.py but no merge_for_ambi.py. do you mean merge_ambi.py ? Sorry, I had another question. I want to run your code on Windows with CPU. Is it possible? If yes, how will the requirments_data.txt file change?
On Sat, Oct 17, 2020 at 4:53 PM Xinpeng Liu notifications@github.com wrote:
Thanks for using and sorry for the late reply! This is a file path issue and we have fixed it in the latest commit.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/DirtyHarryLYL/DJ-RN/issues/7#issuecomment-710953039, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOSYGPYIS64BMYLM3W5VSJTSLGLD3ANCNFSM4SQV3RQA .
A1: Yes. A2: We haven't tested our code on Windows with CPU yet, while we think the requirement doesn't need to change. Sorry for the limited help we could provide. Welcome to ask us if any issue appears.
Hello Excuse me, I'm a beginner. I wanted to connect with one of you who can help me more. I want point clouds of humans and objects.Do you have these point cloud files? thanks
On Mon, Oct 19, 2020 at 9:57 AM Xinpeng Liu notifications@github.com wrote:
A1: Yes. A2: We haven't tested our code on Windows with CPU yet, while we think the requirement doesn't need to change. Sorry for the limited help we could provide. Welcome to ask us if any issue appears.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/DirtyHarryLYL/DJ-RN/issues/7#issuecomment-711645876, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOSYGP7754VRSKPXEZQ6FLLSLPL5VANCNFSM4SQV3RQA .
The point cloud files of humans and objects could take approximately 1TB space, which we couldn't afford to share on cloud. Currently generating the files on a local server is still recommended. We are working on the journal version of this project, and the storage issue would be addressed in the newer version. It might still need some more time for the newer version to be made public.
So can you please guide me which part should be run to get these point clouds ? I need to have the file of these point clouds. please help me. thanks
On Tue, Oct 20, 2020 at 11:36 AM Xinpeng Liu notifications@github.com wrote:
The point cloud files of humans and objects could take approximately 1TB space, which we couldn't afford to share on cloud. Currently generating the files on a local server is still recommended. We are working on the journal version of this project, and the storage issue would be addressed in the newer version, while more time is needed.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/DirtyHarryLYL/DJ-RN/issues/7#issuecomment-712673247, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOSYGP3745PTNPVMCE3HS53SLVAH7ANCNFSM4SQV3RQA .
You could follow this two parts: installation and Data generation to get the point clouds. Follow all the steps included in this two parts, and the point cloud files would be generated.
Hello Sorry to take your time, according to what you said, I run two parts of Installation and Data generation. In the "Data generation" part and in" Generate 3D spatial configuration." step, You take the "--smplx_path <path of the smplx model>" as input but do not use smplx model in the" generate_3D_obj_GT.py" file. You did not use the SMPLify model one at all. How did you get the 3D model? thanks and best regards
The 3D model (smplx model
mentioned above, could be downloaded following step 4 in installation part) is used to get the 3D human joint locations.
https://github.com/DirtyHarryLYL/DJ-RN/blob/41bd02146e38ee34ee5dbb58e0390581c8ac4bf4/script/generate_utils.py#L83
And they are also used by smplify-x to generate 3D human mesh in step 4 of Data generation.
Yes . Now I understand. Only in step 4 there is no line of code. How to get 3D human mesh in step 4 of Data generation https://github.com/DirtyHarryLYL/DJ-RN#data-generation ؟ Does step 5 in Data generation do all this?
On Thu, Oct 22, 2020 at 5:36 PM Xinpeng Liu notifications@github.com wrote:
The 3D model (smplx model mentioned above, could be downloaded following step 4 in installation part https://github.com/DirtyHarryLYL/DJ-RN#installation) is used to get the 3D human joint locations.
https://github.com/DirtyHarryLYL/DJ-RN/blob/41bd02146e38ee34ee5dbb58e0390581c8ac4bf4/script/generate_utils.py#L83 And they are also used by smplify-x to generate 3D human mesh in step 4 of Data generation https://github.com/DirtyHarryLYL/DJ-RN#data-generation.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/DirtyHarryLYL/DJ-RN/issues/7#issuecomment-714518341, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOSYGP743QCX7UYIL4PMUQTSMA35FANCNFSM4SQV3RQA .
You could take reference from the official repo of smplify-x to see how to get 3D human mesh in step 4. And step 5 must be performed after step 1, 2, 3, 4.
hello thank you very much If I have a question in the future, can I bother you?
On Fri, Oct 23, 2020 at 8:04 AM Xinpeng Liu notifications@github.com wrote:
You could take reference from the official repo of smplify-x https://github.com/vchoutas/smplify-x to see how to get 3D human mesh in step 4. And step 5 must be performed after step 1, 2, 3, 4.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/DirtyHarryLYL/DJ-RN/issues/7#issuecomment-714904454, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOSYGPYFU3GIZYUACZK2BZ3SMEBWFANCNFSM4SQV3RQA .
Welcome to bother if the question wasn't covered by our instructions or other issues from this repo and related repos.
hello I want to implement your code. I run "Generate_Ambiguous_HOI.sh". In last line there is "python script/merge_for_ambi.py" but there is not any file "merge_for_ambi.py" in script folder.