Hi ,you honorable author.
You guys did an excellent job on computer vision project. I am so joy when i found your project could be used into some corrections of sports movements, which is the area i am recently working on.
I read some issues as well as i tried to run your programe. Here are some questions i'd like to ask:
(1)If i want to tranverse one set of 2d skeleton points(one input、only one people) into 3d points. I should run the openpose-3d-sandbox right?
(2)Does this programe(openpose-3d-sandbox) try to train a new net?
And about training data:
(1)Is it enough for me to only download the data from H36m? And In what form should i place the data?
I really wish i can start the project at 2017 when you published this work, but i can't. Because i was still in high school at that time. And now there are too many updates in python and otherstuff, as well as the diable of your links. I am trying to overcome this piece by piece.
If you could help. I will be really apreciated!
:）
Hi ,you honorable author.
You guys did an excellent job on computer vision project. I am so joy when i found your project could be used into some corrections of sports movements, which is the area i am recently working on.
I read some issues as well as i tried to run your programe. Here are some questions i'd like to ask:
(1)If i want to tranverse one set of 2d skeleton points(one input、only one people) into 3d points. I should run the openpose-3d-sandbox right?
(2)Does this programe(openpose-3d-sandbox) try to train a new net?
And about training data:
(1)Is it enough for me to only download the data from H36m? And In what form should i place the data?
I really wish i can start the project at 2017 when you published this work, but i can't. Because i was still in high school at that time. And now there are too many updates in python and otherstuff, as well as the diable of your links. I am trying to overcome this piece by piece.
If you could help. I will be really apreciated!
:）