Self-supervised Training of Proposal-based Segmentation via Background Prediction.
Isinsu KatirciogluHelge RhodinVictor ConstantinJörg SpörriMathieu SalzmannPascal FuaPublished in: CoRR (2019)
Keyphrases
- prediction accuracy
- segmentation algorithm
- low depth of field
- foreground and background
- region growing
- medical images
- background pixels
- image segmentation
- moving object segmentation
- segmentation method
- level set
- global thresholding
- prediction model
- foreground objects
- textured objects
- complex background
- adaptive thresholding
- training set
- object regions
- artificial neural networks
- object segmentation
- training examples
- bounding box
- multi layer perceptron
- image analysis
- figure ground segmentation
- multiple objects
- background modeling
- prediction algorithm
- segmentation accuracy
- pixel level
- motion segmentation
- shape prior
- energy function
- supervised learning
- training data