Deep Learning for Automatic Image Captioning in Poor Training Conditions.
Caterina MasottiDanilo CroceRoberto BasiliPublished in: CLiC-it (2017)
Keyphrases
- deep learning
- image features
- input image
- multiscale
- image content
- single image
- image segmentation
- restricted boltzmann machine
- image representation
- image classification
- unsupervised learning
- markov random field
- segmentation method
- supervised learning
- training set
- machine learning
- segmentation algorithm
- energy function
- image regions
- test images
- deep architectures