Semantic Segmentation of Historical Documents via Fully-Convolutional Neural Network.
Ivan GruberMiroslav HlavácMarek HrúzMilos ZeleznýPublished in: SPECOM (2019)
Keyphrases
- semantic segmentation
- convolutional neural network
- historical documents
- handwriting recognition
- face detection
- conditional random fields
- superpixels
- word recognition
- weakly supervised
- scene classification
- document images
- object categories
- pascal voc
- object classes
- object detection
- object class
- object recognition
- multiscale
- image understanding
- neural network
- natural images
- graphical models
- machine learning
- object segmentation
- generative model
- face recognition
- information retrieval