Versatile anomaly detection method for medical images with semi-supervised flow-based generative models.
Hisaichi ShibataShouhei HanaokaYukihiro NomuraTakahiro NakaoIssei SatoDaisuke SatoNaoto HayashiOsamu AbePublished in: Int. J. Comput. Assist. Radiol. Surg. (2021)
Keyphrases
- medical images
- detection method
- generative model
- semi supervised
- semi supervised learning
- detection algorithm
- labeled data
- mixture model
- supervised learning
- deformable models
- face detection
- active learning
- pairwise
- unlabeled data
- discriminative learning
- unsupervised learning
- multi view
- object categories
- pairwise constraints
- co training
- probabilistic model
- metric learning
- prior knowledge
- weakly supervised
- discriminative models
- unlabelled data
- image retrieval
- learning algorithm
- low level features