Learning Deep Structured Multi-Scale Features using Attention-Gated CRFs for Contour Prediction.
Dan XuWanli OuyangXavier Alameda-PinedaElisa RicciXiaogang WangNicu SebePublished in: CoRR (2018)
Keyphrases
- multiscale
- structured output
- unsupervised feature learning
- learning process
- learning algorithm
- max margin learning
- fully connected
- online learning
- prior knowledge
- higher order
- prediction accuracy
- image features
- edge detection
- structured domains
- unsupervised learning
- deep learning
- feature selection
- hidden markov models
- image segmentation
- feature extraction
- training data
- fourier descriptors
- conditional random fields
- feature space
- pairwise
- graphical models
- feature set
- active learning
- co occurrence