Emb-trattunet: a novel edge loss function and transformer-CNN architecture for multi-classes pneumonia infection segmentation in low annotation regimes.
Fares BougourziFadi DornaikaAmir NakibAbdelmalik Taleb-AhmedPublished in: Artif. Intell. Rev. (2024)
Keyphrases
- loss function
- edge detection
- pairwise
- learning to rank
- image segmentation
- segmentation method
- segmentation algorithm
- logistic regression
- support vector
- risk minimization
- level set
- region growing
- reproducing kernel hilbert space
- stochastic gradient descent
- medical images
- active learning
- boosting algorithms
- regularization term
- empirical risk
- semantic annotation
- shape prior
- energy function
- least squares
- feature vectors
- solution path
- boosting framework