zidanmusk / experimenting-with-sort Goto Github PK

I find when the code run in 88 lines,fig.savefig("frameOut/f%d.jpg"%(frame+1),dpi=200),the height of figure become small.Every step in this line can make figure smaller.

what is the specific type of object detector used; F -RCNN, YOLO ? I do not get that

where could i find the file? thanks very much

Hi,
Thanks for this analysis.
I have a doubt that this trackers (bounding boxes would shrink? ) would work when the object is moving away from camera. Can I implement this Sort method with other trackers such as OpenCV based KCF tracker(as it is scale invariant (height and width of bounding boxes of tracker is fixed))?
Could you explain me how new trackers get updated? Are those based on detection or tracking?

The towncenter dataset not showing same result with sort as described in your video when we are using object detectors like SSD or yolo. How can i optimize the sort parameters to get the same?

I am not understanding the detector.py. please be on my side and show me a little of it

Hi,
Does anyone know about .TOP file and how to create it for our own video?

This will not work with any other video or real time video

I am facing the following error when I am done downloading all of the dependancies 👍

File "C:\Python27\lib\site-packages\skimage\io_plugins\pil_plugin.py", line 35, in imread
with open(fname, 'rb') as f:
IOError: [Errno 2] No such file or directory: 'test/Pictures1.jpg'

Please help!
If I have to add some images to any directory, please tell me from where to download those?

can i tracking another object?? ex: car, plane,etc

hello, I saw your code about kalman face tracking，I'm a little confused about the initialization method. Can you answer that for me, thank you very much!

Hello i was implementing kalman object whit sort to a study project that im researching.
Id like to send a signal to a specific kalman tracker object before it is removed from trackers array.
or there is some signal that can i get from inside kalman tracker object before it is erased?

Can anyone guide me as to how I can plug in my custom trained tensorflow object detection into the detector.py? I want the program to take my tensorflow object detected model and do the tracking on it.

Hi abewley ,

I was trying to run SORT on videos.But I found that there are some instances when trackers are not getting deleted(Continuing their linear trajectory) even when there is no detection for few frames.
Code:

Kalman filter update function:

Results seems something like this:

I have used max_age = 2 , min_hits = 5.
It will be a a great help if you could help me out of this issue.

Thanks in advance
Aayush

i got the following error msg

Traceback (most recent call last):
File "main.py", line 110, in
main()
File "main.py", line 56, in main
img = io.imread(fn)
File "/usr/local/lib/python2.7/dist-packages/skimage/io/_io.py", line 61, in imread
img = call_plugin('imread', fname, plugin=plugin, **plugin_args)
File "/usr/local/lib/python2.7/dist-packages/skimage/io/manage_plugins.py", line 211, in call_plugin
return func(*args, **kwargs)
File "/usr/local/lib/python2.7/dist-packages/skimage/io/_plugins/pil_plugin.py", line 35, in imread
with open(fname, 'rb') as f:
IOError: [Errno 2] No such file or directory: 'test/Pictures1.jpg'

I test this network for another video/image.
But I think that this sample is only for TownCentreXVID.avi
right ?