Structured Prediction with Output Embeddings for Semantic Image Annotation.
Ariadna QuattoniArnau RamisaPranava Swaroop MadhyasthaEdgar Simo-SerraFrancesc Moreno-NoguerPublished in: HLT-NAACL (2016)
Keyphrases
- image annotation
- structured prediction
- semantic concepts
- automatic image annotation
- low level features
- image understanding
- maximum margin
- multi modal
- visual features
- conditional random fields
- multi label
- visual concepts
- image content
- image retrieval
- markov networks
- efficient learning
- high level
- latent variables
- low level
- max margin
- semantic information
- natural language
- convex optimization
- approximate inference
- information retrieval
- higher order
- multi label classification
- relevance feedback
- multi class
- feature extraction
- computer vision