Tutorial for how to use FixedSplitter

[ribenamaplesyrup]

This tutorial shows how to split a dataset using predefined imagesets as specified in issue #585. The dataset used within the tutorial is Pascal VOC 2012. This tutorial will only work using the latest icedata (the one via PyPI won't work as specified in "AttributeError: module 'icedata' has no attribute 'voc' #84") with my proposed changes in PR "Add voc data functions #91".

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[codecov]

Codecov Report

Merging #594 (3a95661) into master (6cbd6d1) will increase coverage by 0.66%.
The diff coverage is 98.87%.

@@            Coverage Diff             @@
##           master     #594      +/-   ##
==========================================
+ Coverage   88.15%   88.82%   +0.66%     
==========================================
  Files         113      114       +1     
  Lines        2694     2854     +160     
==========================================
+ Hits         2375     2535     +160     
  Misses        319      319              

Flag	Coverage Δ
unittests	88.82% <98.87%> (+0.66%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
icevision/models/utils.py	64.86% <ø> (ø)
icevision/visualize/show_data.py	78.12% <50.00%> (ø)
icevision/models/interpretation.py	99.15% <99.15%> (ø)
icevision/models/efficientdet/show_results.py	95.83% <100.00%> (+5.83%)	⬆️
...odels/torchvision_models/faster_rcnn/prediction.py	100.00% <100.00%> (+3.44%)	⬆️
...els/torchvision_models/faster_rcnn/show_results.py	100.00% <100.00%> (ø)
...s/torchvision_models/keypoint_rcnn/show_results.py	100.00% <100.00%> (ø)
...odels/torchvision_models/mask_rcnn/show_results.py	92.30% <100.00%> (+2.30%)	⬆️
...odels/torchvision_models/retinanet/show_results.py	100.00% <100.00%> (ø)
icevision/utils/imageio.py	80.39% <100.00%> (+5.39%)	⬆️
... and 2 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 6cbd6d1...068c287. Read the comment docs.

[lgvaz]

Alright, perfect!

This tutorial will be very helpful! I'll make a subsequent commit to add this to the docs

Thanks a lot!