Latent Space Regularization for Unsupervised Domain Adaptation in Semantic Segmentation.
Francesco BarbatoMarco ToldoUmberto MichieliPietro ZanuttighPublished in: CoRR (2021)
Keyphrases
- semantic segmentation
- latent space
- generative model
- conditional random fields
- gaussian process
- low dimensional
- superpixels
- object categories
- feature space
- dimensionality reduction
- scene classification
- high dimensional
- transfer learning
- multiscale
- distance metric
- object classes
- probabilistic model
- prior information
- image processing
- data sets
- model selection
- em algorithm
- natural images
- semi supervised