Handling Missing Annotations for Semantic Segmentation with Deep ConvNets.
Olivier PetitNicolas ThomeArnaud CharnozAlexandre HostettlerLuc SolerPublished in: DLMIA/ML-CDS@MICCAI (2018)
Keyphrases
- semantic segmentation
- feed forward
- street scenes
- conditional random fields
- scene classification
- superpixels
- label transfer
- weakly supervised
- object categories
- image annotation
- back propagation
- artificial neural networks
- image understanding
- neural network
- object classes
- object class
- hidden markov models
- probabilistic model
- pascal voc
- object recognition
- image classification
- input image
- bounding box
- pairwise
- annotated images
- bayesian networks