Towards optimal vlad for human action recognition from still images.
Lei ZhangXiantong ZhenJiqing HanPublished in: ICASSP (2016)
Keyphrases
- image analysis
- ground truth
- image classification
- image data
- image retrieval
- image database
- three dimensional
- image matching
- object recognition
- image features
- optimal solution
- rigid body
- image collections
- region of interest
- image set
- illumination conditions
- lighting conditions
- image understanding
- edge detection
- image annotation
- closed form
- segmentation method
- multiple images
- original images
- image search
- segmentation algorithm
- spatial information
- feature points
- input image
- image registration
- dynamic programming
- feature extraction