Enlarging the discriminability of bag-of-words representations with deep convolutional features.
Daniel MangerDieter WillersinnPublished in: IPTA (2017)
Keyphrases
- bag of words
- unsupervised feature learning
- visual words
- image classification
- textual features
- deep learning
- image representation
- action recognition
- image features
- bag of words models
- text classification
- text classification tasks
- bag of features
- co occurrence
- n gram
- feature representations
- spatial pyramid
- deep belief networks
- document representation
- spatial pyramid matching
- feature set
- feature extraction
- data mining
- spatial layout
- neural network
- text documents
- visual vocabulary
- object categorization
- late fusion
- image segmentation
- unsupervised learning
- feature space
- visual codebook
- multiscale