[FIX] Enable VOC dataset in OTX

[supersoob]

Issue

• VOC dataset is not properly supported in OTX
• Cannot exactly detect the train/val subset specified in VOC dataset

Root Cause

• When get_train_dataset from datumaro format, it checks key includes any "train" str. Thus, it first bring the subset dataset named "trainval", not "train"
• When get_val_dataset from datumaro format, it checks key includes any "val" str. Thus, it first bring the subset dataset named "trainval", not "val"
• Only if masks annotation exists, dataset item is valid. but it included the shape with no annotation.

Solution

• Resolved by checking whole words, "train" and "val" first

[sungmanc]

Nice work :)

[kprokofi]

Will it also fix COCO dataset bug, when it uses the same "trainval" instead of "train" and "val" (for instance segmentation)?

[sungmanc]

Will it also fix COCO dataset bug, when it uses the same "trainval" instead of "train" and "val" (for instance segmentation)?

Yes, if user prepare "instances_trainval.json", then we can avoid including this annotation to train annotation.

[codecov-commenter]

Codecov Report

Patch coverage: 100.00% and project coverage change: -0.03 ⚠️

Comparison is base (872f119) 80.55% compared to head (3b39abf) 80.53%.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #1862      +/-   ##
===========================================
- Coverage    80.55%   80.53%   -0.03%     
===========================================
  Files          481      481              
  Lines        32986    32995       +9     
===========================================
- Hits         26573    26572       -1     
- Misses        6413     6423      +10     

Impacted Files	Coverage Δ
.../core/data/adapter/segmentation_dataset_adapter.py	94.01% <100.00%> (+0.05%)	⬆️
otx/core/data/manager/dataset_manager.py	97.18% <100.00%> (-2.82%)	⬇️
...dapters/openvino/model_wrappers/openvino_models.py	74.00% <0.00%> (-3.00%)	⬇️
otx/algorithms/detection/tasks/inference.py	71.16% <0.00%> (-1.88%)	⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.