Google Colab demo with video input

[sharmaabhishekk]

Thank you for this repository.

I am trying to get a AlphaPose+PoseFlow workflow on videos to run on Google Colab. The AlphaPose part is fine - I'm getting the alphapose-results.json.

However, how do I run PoseFlow on it (since I don't have a image directory -just a video - and image_dir is a necessary argument for tracker-general.py)

My notebook is here.

[jpbranson]

The only solution I could find was to extract all the video frames into a directory - run AlphaPose on that directory, then use the result to run PoseFlow.

Also, you may already know this, but I think PoseFlow requires Python 2.7. I had to create separate Google Colab notebooks - one for AlphaPose with Python 3, and one for PoseFlow with Python 2.7. I manually moved the alphapose_results.json from the former to the latter. I may be missing something though, I'm new to Python

[sharmaabhishekk]

Yup, ended up doing that as well. I guess I'll leave this open for now in case the authors have a better solution to this.

[FCrSTATS]

I just changed xrange to range in the general_tracker.py file and it seemed to work in python, but maybe it impacted the results. Would this fix provide clean results?

[jpbranson]

@FCrSTATS It would make sense to provide the same results. We could try running it on the same clip to test?

[neeek2303]

@sharmaabhishekk hi! Could you please share your code with PoseFlow part?