Two data selection techniques (GMM visual similarity and bounding-box diversity) reduce required weakly labeled images by up to 100x on Open Images and 20x on Cityscapes while maintaining semantic segmentation performance.
Taskonomy: Disentangling task transfer learning,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Data Selection for training Semantic Segmentation CNNs with cross-dataset weak supervision
Two data selection techniques (GMM visual similarity and bounding-box diversity) reduce required weakly labeled images by up to 100x on Open Images and 20x on Cityscapes while maintaining semantic segmentation performance.