sklearn version compatibility

[OmerShubi]

Describe the bug

/home/shubi/.conda/envs/cn_client_pd_fix3/lib/python2.7/site-packages/sklearn/base.py:253: UserWarning: Trying to unpickle estimator LinearRegression from version pre-0.18 when using version 0.20.3. This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)

To Reproduce
Steps to reproduce the behavior:
python start_local.py -d "/home/shubi/Desktop/experiment_23_09_2019"
Play around in GUI

Expected behavior
No warning

[OmerShubi]

Log from the start_local, after pressing seek in GUI.

Seeking: /home/shubi/Desktop/experiment_23_09_2019/cam_099/captured_images/2017_04_22/1492826400.0_2017_04_22_02_00_00_0.mat

Searching for ocam path: /home/shubi/PycharmProjects/cameranetwork/CameraNetwork/../data/calibration/4103098535/2017_03_22/ocamcalib.pkl

/home/shubi/.conda/envs/cn_client_pd_fix3/lib/python2.7/site-packages/sklearn/base.py:253: 
UserWarning: Trying to unpickle estimator PolynomialFeatures from version pre-0.18 when using version 0.20.3. 
This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)

/home/shubi/.conda/envs/cn_client_pd_fix3/lib/python2.7/site-packages/sklearn/base.py:253: 
UserWarning: Trying to unpickle estimator LinearRegression from version pre-0.18 when using version 0.20.3. 
This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)

/home/shubi/.conda/envs/cn_client_pd_fix3/lib/python2.7/site-packages/sklearn/base.py:253: 
UserWarning: Trying to unpickle estimator RANSACRegressor from version pre-0.18 when using version 0.20.3. 
This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)

/home/shubi/.conda/envs/cn_client_pd_fix3/lib/python2.7/site-packages/sklearn/base.py:253: 
UserWarning: Trying to unpickle estimator Pipeline from version pre-0.18 when using version 0.20.3. 
This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)

[OmerShubi]

This is attributed to the vignetting.pkl. Tried to open models with sklearn version 0.20.3 and they seem to be okay.