Zero-Shot Learning of Language Models for Describing Human Actions Based on Semantic Compositionality of Actions.
Hideki AsohIchiro KobayashiPublished in: PACLIC (2014)
Keyphrases
- human actions
- language model
- action recognition
- spatio temporal
- language modeling
- human motion
- human activities
- action sequences
- space time
- probabilistic model
- n gram
- document retrieval
- language modelling
- recognizing actions
- recognizing human actions
- visual features
- query expansion
- statistical language models
- retrieval model
- activity recognition
- information retrieval
- video sequences
- test collection
- recognition of human actions
- action models
- action primitives
- relevance model
- bag of words
- context sensitive
- image search
- smoothing methods
- moving objects
- machine learning
- three dimensional
- human computer interaction
- semantic information
- image retrieval
- bayesian networks
- image classification
- semantic similarity
- language models for information retrieval