zhengyiluo / embodiedpose Goto Github PK
View Code? Open in Web Editor NEWOfficial implementation of NeurIPS 2022 paper: "Embodied Scene-aware Human Pose Estimation".
Home Page: https://zhengyiluo.github.io/projects/embodied_pose/
Official implementation of NeurIPS 2022 paper: "Embodied Scene-aware Human Pose Estimation".
Home Page: https://zhengyiluo.github.io/projects/embodied_pose/
Hi, thanks for releasing your code!
I am trying to run eval_scene.py
with the default values and I get stuck in kin_net_humor_res.py
asking me for alignment files with the form: f'{prox_path}/alignment/{scene_name}.npz'
(e.g., prox/qualitative/alignment/MPH8.npz) .
I cannot find these files in PROX dataset. Where can I get these files?
Thanks!
Hi, I used the script data_process/process_humor_split.py
to generate the same data as the one provided in sample_data/thirdeye_anns_prox_overlap_no_clip.pkl
for PROX "N0Sofa_00145_01" sequence. I have the same data fields as the file provided in the repo, however, when I run the method I don't have the same results. When using sample_data/thirdeye_anns_prox_overlap_no_clip.pkl
data, it works ok but when I use the data generated by process_humor_split.py
the result is not good.
Can you provide the code used for generating sample_data/thirdeye_anns_prox_overlap_no_clip.pkl
?
Thanks. Best,
Hi,
Is is possible to run you method in videos-in-the-wild and, if not, how should I process my own 2D keypoints estimations as to get the 3D predictions?
Do I first need to process the data using python uhc/data_process/process_smpl_data.py
?
Thanks!
Hi, thank you for your great work. I have some questions about how to simplify the scene.
In the paper, you say you simplify the scene semi-automatically by first selecting key points in the mesh vertices and then creating rectangular cuboids and cylinders.
I want to know how to select the key points and create the cuboids. Are these operations done in MeshLab or Unity3D? I would appreciate it if you can give me some more instructions about the procedure.
Looking forward to your reply!
Hi,
When I run the script process_smpl_data.py
I get the following error.
TypeError: smpl_to_qpose() missing 1 required positional argument: 'mj_model'
at EmbodiedPose/uhc/data_process/process_smpl_data.py", line 142, in smpl_2_entry.
Thanks
Thanks for your great work!
When will the training code be released?
Best wishes.
error:
Failed to load XML from file: C:\Users\admin\AppData\Local\Temp\tmp5h7qhqh9.xml. mj_loadXML error: b"XML Error: Schema violation: unrecognized attribute: 'name'\n\nElement 'exclude', line 264, column 5\n"
i use dance.mp4 as the wild video to run it. it run step1 and step2 is good,
but when run step3: python scripts/eval_scene.py --cfg tcn_voxel_4_5 --epoch -1 --data sample_data/wild_processed.pkl
show me error, these are nothing changed.
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.