bugfix in video demo and feature in visualization#5
bugfix in video demo and feature in visualization#5TheTesla wants to merge 10 commits intopulkitag:masterfrom TheTesla:master
Conversation
…of pose from last prediction to initialize next prediction, optional fixed scale (disable netScale); current Problems: runaway
…s pose; initialisation is weighted with loopfactor
|
I will have a look at it this week. Sorry for the delay in response. |
… reuse for same person and csv person number; added frame number to csv; parameters about human detection
|
Hi Stefan, A couple of things before I merge this:
|
|
I am doing the work to prepare my contributions for your needs. Should I use 4 or 8 space for indentation? Will you provide the code for training? |
|
4 space. Hoping to release the training code in Dec. On Oct 14, 2016 3:06 PM, "Stefan Helmert" notifications@github.com wrote:
|
bugfix: autodetect opencv major version - variable is string must be converted in int;
feature: RGB->BGR convertion optional in plot_pose_stickmodel_cv2mat()