Bayesian Deep Learning with Monte Carlo Dropout for Qualification of Semantic Segmentation.
Clément DechesnePierre LassalleSébastien LefèvrePublished in: IGARSS (2021)
Keyphrases
- monte carlo
- deep learning
- weakly supervised
- superpixels
- relation extraction
- object class
- topic models
- semi supervised
- markov chain
- unsupervised learning
- named entities
- maximum likelihood
- multiscale
- bayesian networks
- conditional random fields
- machine learning
- object detectors
- bounding box
- segmentation algorithm
- viewpoint