Using Mise-En-Scène Visual Features based on MPEG-7 and Deep Learning for Movie Recommendation.
Yashar DeldjooMassimo QuadranaMehdi ElahiPaolo CremonesiPublished in: CoRR (2017)
Keyphrases
- deep learning
- visual features
- movie recommendation
- visual descriptors
- visual content
- recommender systems
- visual information
- image classification
- unsupervised learning
- low level
- machine learning
- low level features
- collaborative filtering
- image retrieval
- keywords
- content based filtering
- image annotation
- weakly supervised
- mental models
- video sequences
- feature selection
- saliency map
- color histogram
- user profiles
- image representation
- high level