drmahdirezaei / deepsocial Goto Github PK

Thank you uploading this very interesting work! How did you select the calibration points;

calibration = [[180,162],[618,0],[552,540],[682,464]]

for the Oxford Town Centre dataset?

hi, what method is used to realize pedestrian distance visualization ，I tried it with python but the effect is not good

calibration = [[180,162],[618,0],[552,540],[682,464]]
这个为什么这样取值啊，冒昧问一下。

When I run the code on VSCode ,I got error as "ImportError: bad magic number in 'deepsocial': b'B\r\r\n'" and I coundn't deal with it. May I ask for the original code of deppsocial.py or any other solutions? Thank you very much

from deepsocial import *
however error occur:ImportError: bad magic number in 'deepsocial': b'B\r\r\n'
i checked it is bacause the version of python of colab is not the same as authors

Hello! I'm very interested in your excellent work!
I wanna program the jupyter notebook as a project. After I performed the last step, I got log as below. I wanna know if I installed the mismatch version of opencv. I installed opencv3.4.4.

_crowdMap, crowdMap = Apply_crowdMap(centroid_dict, image, _crowdMap)
File "deepsocial.py", line 237, in Apply_crowdMap
TypeError: Layout of the output array img is incompatible with cv::Mat (step[ndims-1] != elemsize or step[1] != elemsize*nchannels)

Hello, Dr.MahdiRezaei!
Thanks for sharing your method. I am wondering whether there is people's 3d groundtruth location labels for your provided Oxford Town Dataset.