search

Engineering-Course / LIP_JPPNet

Code repository for Joint Body Parsing & Pose Estimation Network, T-PAMI 2018
MIT License
341 stars 87 forks source link

Pose estimation output interpretation #6

Closed achalshah20 closed 6 years ago

achalshah20 commented 6 years ago

Hi, I just run inference on some random images using evaluate_pose_JPPNet-s2.py. I got output file with some numbers. Can you please tell me what these numbers are?

P.S. If these are pixel locations then please let me know if there is a pattern to connect these pixel to draw lines like in your paper.

Is there any way I can visualize this like you did in your paper? Thanks !

Engineering-Course commented 6 years ago

Each row of the output file has 32 numbers for the positions of 16 key points. (x1, y1, x2, y2....).

achalshah20 commented 6 years ago

What are these keypoints? Is there any mapping to order of x,y indices?

Can you share the visualization code which you used to show these 16 keypoints as cuts shown in paper. Thanks!

achalshah20 commented 6 years ago

So, I wrote the modified version of vis_annotation.py file(from dataset) which is enough to visualize these keypoints. Thanks!

chnanc001 commented 6 years ago

@achalshah20 to which vis_annotation.py file are you referring?