Calibrating the Dice loss to handle neural network overconfidence for biomedical image segmentation.
Michael YeungLeonardo RundoYang NanEvis SalaCarola-Bibiane SchönliebGuang YangPublished in: CoRR (2021)
Keyphrases
- neural network
- image segmentation
- back propagation
- multiscale
- graph cuts
- image analysis
- fuzzy logic
- artificial neural networks
- level set method
- markov random field
- active contours
- computer vision
- pattern recognition
- boundary detection
- machine learning
- region growing
- neural network model
- medical imaging
- fault diagnosis
- self organizing maps
- biomedical data
- contour detection
- biomedical images
- texture segmentation
- expert systems
- training process
- neural nets
- bp neural network
- shape prior
- recurrent neural networks
- multilayer perceptron
- segmentation algorithm
- deformable models
- neural network is trained
- level set
- method for image segmentation
- segmentation method
- camera calibration
- feed forward neural networks
- gray level
- image processing
- multi layer perceptron
- radial basis function
- fuzzy neural network
- feed forward
- prediction model