Tf-pose-estimation: License type

[jpizarrom]

Hi,
@ildoonet thank for sharing.

It will be very helpful to know if you had used code of [1][5][6], their license is "ACADEMIC OR NON-PROFIT ORGANIZATION NONCOMMERCIAL RESEARCH USE ONLY" [2], and your project is under "Apache License 2.0" [3].

I am asking because I want to port your project [4] to Android Using SNPE SDK, therefore I want to know which l license should I use.

[1] https://github.com/CMU-Perceptual-Computing-Lab/openpose
[2] https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/LICENSE#L3
[3] https://github.com/ildoonet/tf-pose-estimation/blob/master/LICENSE
[4] https://github.com/ildoonet/tf-pose-estimation
[5] https://github.com/ZheC/Realtime_Multi-Person_Pose_Estimation
[6] https://github.com/CMU-Perceptual-Computing-Lab/caffe_rtpose

[jpizarrom]

In [1] and [2] they changed the licenses from GPL to other.

openpose never was GPL [3]

[1] https://github.com/CMU-Perceptual-Computing-Lab/caffe_rtpose/commit/122d9dbb9a72150e777af1fafaf29f99617a3311#diff-9879d6db96fd29134fc802214163b95a
[2] https://github.com/ZheC/Realtime_Multi-Person_Pose_Estimation/commit/e19a67d1df9b5d5c4994a252b3e37a15ba6f6ae6#diff-9879d6db96fd29134fc802214163b95a
[3] https://github.com/CMU-Perceptual-Computing-Lab/openpose/commits/master/LICENSE

[ildoonet]

I didn't used [1]'s codes but the model named cmu is copied from theirs.

[5]'s codes were very helpful, but I implemented in a different way and in a different language. But it is somewhat similar.

I didn't used [6]'s codes.

[jpizarrom]

@ildoonet Thanks for sharing

[s101010tw]

Hi @jpizarrom,
Sorry, I have one question about Android SNPE SDK,
Did you successfully port the project [4] to Android using SNPE SDK?
If yes, how about the performance (FPS)?

Thanks

[kristijanbartol]

Is it therefore OK to use the mobilethin model for commercial use in an Android app?

[kartikeyporwal]

I didn't used [1]'s codes but the model named cmu is copied from theirs.

[5]'s codes were very helpful, but I implemented in a different way and in a different language. But it is somewhat similar.

I didn't used [6]'s codes.

Hi, nice work!!

Since you added license type as commercial but used CMU's Caffe Model and converted to Tensorflow model. But, we can't use their pretrained model which follows the same license as their code, this is the issue, I'm referring to.

Converting CMU's pretrained model is a derivative of their work, so it's owned by CMU based on the license of openpose. Could you look this issue from OpenPose repo for more detail?

Though I'm uncertain about your mobilenet models.

So, this or any other implementation of Openpose can't be used commercially for free!

What do you think about this?