Feature Selection for HOG Descriptor Based on Greedy Algorithm.
Yonghwa ChoiSungmoon JeongMinho LeePublished in: ICONIP (3) (2013)
Keyphrases
- greedy algorithm
- feature selection
- action recognition
- feature set
- histogram of oriented gradients
- histograms of oriented gradients
- feature extraction
- object detection
- greedy algorithms
- dynamic programming
- human detection
- worst case
- objective function
- multi class
- bag of words
- feature space
- pedestrian detection
- feature descriptors
- set cover
- influence maximization
- machine learning
- local binary pattern
- keypoints
- multi task
- text classification
- influence spread
- approximation guarantees
- submodular functions
- approximation ratio
- greedy heuristic
- scheduling problem
- special case
- feature vectors
- multiscale
- greedy strategy
- learning algorithm