Augusto Escalante

Results 24 comments of Augusto Escalante

Hi @MMathisLab Sorry, I just added the DLC version. It is actually the one COLAB currently uses (2.2.1). I did not test outside COLAB, for the lack of a proper...

Hi @jeylau , I just ran both with use_shelve=False and True. Attached you can see the progression of RAM use. I took the screenshot a couple of seconds before the...

It does. Actually, those videos are already analyzed (as you can read in the cell output) but with auto_track=False. Now we just want to get the tracks (remember our previous...

You are right, I should have phrased the issue correctly, sorry!

The issue appears already with `convert_detections2tracklets` : ![Screenshot from 2022-05-03 16-09-56](https://user-images.githubusercontent.com/24290290/166469904-670b524b-4d5c-4c70-b1fd-84536349ce74.png)

Hi @jeylau, that sounds great. Let me know when its available for COLAB use! Thanks!

Ok. Mixed feelings: When runnning directly `analyze_videos` with `auto_track=True`, `use_shelve=True` is apparently not passed to `convert_detections2tracklets` and RAM is used as much as before: ![Analyze](https://user-images.githubusercontent.com/24290290/166743920-5470b7a2-6457-4881-8d45-cb0b59fb0fbb.png) When running `convert_detections2tracklets` RAM is...

Just to be clear, you mean use_shelve=True should have been passed directly to convert_detections2tracklets from analyze_videos? Here are all the files I got for this particular video: ![Screenshot from 2022-05-04...

Hi @roomrys Thanks a lot for your help. Then it might have been our mistake: we thought when ticking the Grayscale option in the GUI the conversion would be done...