muelea / tuch Goto Github PK
View Code? Open in Web Editor NEWCVPR 2021 - Official code repository for the paper: On Self-Contact and Human Pose.
Home Page: https://tuch.is.tue.mpg.de
License: Other
CVPR 2021 - Official code repository for the paper: On Self-Contact and Human Pose.
Home Page: https://tuch.is.tue.mpg.de
License: Other
Hi, @muelea ! Thanks for your great work. I found the mtp dataset only provides smpl pose and shape parameter, missing the smpl transition parameter and camera parameters. I want to render the smpl ground truth on the image. It would be very much appreciated if you have any advice on this.
Hi @muelea ,
I visit the DeepFashion Website and found that there are several benchmarks, each with different zips for images. I download one of the zips, but after extraction, the folder is like this:
DeepFashion/WOMEN/Tees_Tanks/id_00007979/04_4_full.jpg
DeepFashion/WOMEN/Tees_Tanks/id_00007979/04_7_additional.jpg
DeepFashion/WOMEN/Tees_Tanks/id_00007981/
DeepFashion/WOMEN/Tees_Tanks/id_00007981/03_1_front.jpg
DeepFashion/WOMEN/Tees_Tanks/id_00007981/03_3_back.jpg
DeepFashion/WOMEN/Tees_Tanks/id_00007981/03_7_additional.jpg
rather than the structure that all images in one folder:
${DF_ROOT}
|-- images
|-- img_00000001.jpg
|-- img_00000002.jpg
I guess I downloaded the wrong subset of dataset of DF.
Could you give more details on which benchmark (or even better, which zip :)) to download? Thanks in advance.
Hi,
I have a quick question regarding the MTP dataset. I noticed that the left_hand_pose
and right_hand_pose
are 6-dimensional PCA coefficients for 3DCP Scan
and 3DCP Mocap
. However, when visualizing for AGORA
, I noticed the vectors are 12 dimensional.
Could you please tell me how to interpret this specifically for AGORA? I noticed that in several cases, the latter 6 values were all zeros, but this is not consistent for all cases. I just wanted to make sure I was not missing anything.
Also, thank you for the great work and for providing the code and data! :)
Hi @muelea , Thanks for sharing your program with us, it' clean and helpful.
When running the SMPLify-DC demo
command, I found that the pkl files are missing:
python demo_smplify_dc.py --name smplify_dc --log_dir out/demo_smplify_dc --ds_names dsc_df \
--num_smplify_iters 100
Traceback (most recent call last):
File "demo_smplify_dc.py", line 183, in <module>
main(options)
File "demo_smplify_dc.py", line 45, in main
dataset = BaseDataset(options, options.ds_names[0], use_augmentation=False)
File "/data_1/xxx/projects/github/tuch/tuch/datasets/base_dataset.py", line 67, in __init__
self.classes = pickle.load(open(osp.join(config.DSC_ROOT, 'classes.pkl'), 'rb'))
FileNotFoundError: [Errno 2] No such file or directory: '/data_1/xxx/datasets/DSC/classes.pkl'
Could you please upload "classes.pkl" and "ContactSigSMPL.pkl" ? Thanks
Hi @muelea ,
I am trying to find the checkpoints which are being referenced in these lines in the config but can't seem to find them anywhere. Could you please provide the links where I could download these from?
I really appreciate any help you can provide. Thanks!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.