Empty mask fix

[mayrajeo]

Related to #913, fixes to the following:

• Force empty mask annotations to have valid shape with quick hack.
  • This was breaking COCOMetric(metric_type=COCOMetricType.mask)
• Avoid trying to concatenate empty masks
  • This was breaking a lot of things
• Just to be sure add check for empty self._keep_mask list in albumentations_adapter/_filter_attribute
  • This might not be needed anymore. Seems that original problem was related to faulty masks but at least I found some other bugs to squish

[codecov]

Codecov Report

Merging #933 (be6e327) into master (0dce47c) will increase coverage by 0.04%.
The diff coverage is 71.42%.

@@            Coverage Diff             @@
##           master     #933      +/-   ##
==========================================
+ Coverage   85.62%   85.67%   +0.04%     
==========================================
  Files         251      251              
  Lines        5455     5487      +32     
==========================================
+ Hits         4671     4701      +30     
- Misses        784      786       +2     

Flag	Coverage Δ
unittests	85.67% <71.42%> (+0.04%)	⬆️

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

Impacted Files	Coverage Δ
icevision/core/mask.py	81.75% <66.66%> (-0.77%)	⬇️
...sion/tfms/albumentations/albumentations_adapter.py	96.15% <100.00%> (+0.10%)	⬆️
icevision/core/tasks.py	100.00% <0.00%> (ø)
icevision/metrics/metric.py	100.00% <0.00%> (ø)
icevision/models/mmdet/fastai/callbacks.py	100.00% <0.00%> (ø)
icevision/models/torchvision/fastai/callbacks.py	100.00% <0.00%> (ø)
...sion/models/torchvision/lightning/model_adapter.py	100.00% <0.00%> (ø)
icevision/core/record_components.py	81.57% <0.00%> (+0.05%)	⬆️
icevision/models/mmdet/lightning/model_adapter.py	96.96% <0.00%> (+0.09%)	⬆️
... and 7 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 0dce47c...be6e327. Read the comment docs.

[ai-fast-track]

Thanks a lot for your contribution @jaeeolma!

Sorry for the long delay.