Context-aware horror video scene recognition via cost-sensitive sparse coding.
Xinmiao DingBing LiWeiming HuWeihua XiongZhenchong WangPublished in: ICPR (2012)
Keyphrases
- context aware
- cost sensitive
- sparse coding
- image representation
- sparse representation
- natural images
- multi class
- image classification
- linear combination
- generative model
- unsupervised learning
- active learning
- visual words
- scene classification
- feature space
- multiscale
- multi label
- support vector machine
- high dimensional
- object detection
- small number
- higher order
- denoising
- image processing
- probabilistic latent semantic analysis