sgoldenlab / simba

SimBA (Simple Behavioral Analysis), a pipeline and GUI for developing supervised behavioral classifiers
https://simba-uw-tf-dev.readthedocs.io/
GNU General Public License v3.0
289 stars 141 forks source link

"Visualise pose-estimation" tool not working #266

Closed alisabak closed 1 year ago

alisabak commented 1 year ago

Dear SimBA team,

I am trying to use a tool "Visualise pose-estimation", but it opens an empty window (screenshot below) and gives me the following error:

Exception in Tkinter callback Traceback (most recent call last): File "C:\Users\Alisa\anaconda3\envs\simba\lib\tkinter\__init__.py", line 1705, in __call__ return self.func(*args) File "C:\Users\Alisa\anaconda3\envs\simba\lib\site-packages\simba\ui\pop_ups\visualize_pose_in_dir_pop_up.py", line 18, in __init__ self.circle_size = Entry_Box(settings_frame, 'Circle size', 0, validation='numeric', labelwidth=20) TypeError: __init__() got multiple values for argument 'labelwidth'

image

I would greatly appreciate your help!

Desktop (please complete the following information):

sronilsson commented 1 year ago

Thank you @alisabak for reporting, we appreciate it! If you upgrade simba to version 1.63.0 with pip install simba-uw-tf-dev --upgrade or pip install simba-uw-tf-dev==1.63.0, how does it look on your end?

alisabak commented 1 year ago

Thank you for such a quick fix! It works now.

Would it be possible to visualize tracking for individual videos? I takes a lot of time to generate videos for the whole project.

sronilsson commented 1 year ago

Thanks @alisabak - if you temporarily move some files out of the input directory and only keep the ones you want to visualize in the input folder, you will only get some videos.

But yes, looking at this method, it looks a little dated and slow, no options to multi-process... I will see if I can update it and let you.

sronilsson commented 1 year ago

If you update SimBA again you should see this:

image

You still can't visualize a single video in this method (it has to be a directory). But: if you select a higher number of cores in the Core count dropdown it will go much faster (follow progress in main OS terminal).

Sometimes users just want to check if the pose-estimation looks good following smoothing/interpolation using this method, and they may not need the entire video. If that's the case, use the Video output sample sizes (s) dropdown. Here, if you select e.g., 100, then only the first 100s of each video will be created and things will be a little faster.. Let us know if somethings missing Thanks again

alisabak commented 1 year ago

Thank you @sronilsson, that is exactly what I needed!