Comments (9)
I found that I should use the miss images to compare with, but there are still slightly different from yours.
Does that affect the training result?
Is that differences makes you perform better?
Here's the comparison:
https://imgur.com/a/v3XucJk
from pytorch_realtime_multi-person_pose_estimation.
Training steps of the official repo: https://github.com/ZheC/Realtime_Multi-Person_Pose_Estimation will do that. :)
from pytorch_realtime_multi-person_pose_estimation.
Thanks, I will try.
from pytorch_realtime_multi-person_pose_estimation.
Another issue I found is why the mask images I generated with genCOCOMask.m are different from yours?
Here's the comparison:
https://imgur.com/a/LBiC2OC
from pytorch_realtime_multi-person_pose_estimation.
I'm not sure if the processing you generated mask is right, but that mask is for the unlabeled person, and the mask you made seems like the ground truth semantic label.
from pytorch_realtime_multi-person_pose_estimation.
I just ran the training steps from the official repo as you told:
- Run getANNO.m in matlab to convert the annotation format from json to mat in dataset/COCO/mat/.
- Run genCOCOMask.m in matlab to obatin the mask images for unlabeled person. You can use 'parfor' in matlab to speed up the code.
What changes should be made to generate mask images like yours?
from pytorch_realtime_multi-person_pose_estimation.
Hi, @xind. I just generated the mask with the same processing as you as I can remember. I don't think there are many differences for the final result caused by the mask. :)
from pytorch_realtime_multi-person_pose_estimation.
@tensorboy Got it! Thanks for your replies.
from pytorch_realtime_multi-person_pose_estimation.
I want to train with train2017 and I've tried using coco.annToMask() to get mask but the output are different from yours.
How can I get mask for train2017?
Hi, I want train coco2017 like you, could you be kind to share me the changes on this repo you have done?
from pytorch_realtime_multi-person_pose_estimation.
Related Issues (20)
- no results on own datasets test HOT 3
- mystical memory cost issues
- Hello,where is “slim”? HOT 1
- Some questions about training?
- ImportError: DLL load failed while importing _pafprocess HOT 9
- model from 'rtpose' link cannot archieved a good performance
- Runs very slow on ubuntu 20.04
- 去掉隐藏点训练为啥效果很差
- Cannot open the dropbox HOT 1
- Same results as openpose demo?
- The following error occurred when I tested my trained model. HOT 4
- Is there any post-processing code for C++
- Loss nan
- In the demo code, update the lines to load the model as below:
- how to get multi person heatmaps/confidence maps
- How can the model achieve real-time?
- 您好,请问您有保存最后的optimizer的dict吗
- who can share pretrained model?
- Can I get openpose.pkl file?
- How to training base on MobileNetV2?
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from pytorch_realtime_multi-person_pose_estimation.