Toward Always-On Mobile Object Detection: Energy Versus Performance Tradeoffs for Embedded HOG Feature Extraction.
Alex Omid-ZohoorChristopher YoungDavid TaBoris MurmannPublished in: IEEE Trans. Circuits Syst. Video Technol. (2018)
Keyphrases
- object detection
- feature extraction
- pedestrian detection
- object class
- face detection
- multi class
- computer vision
- object categories
- preprocessing
- mobile devices
- mobile phone
- object recognition
- feature vectors
- wavelet transform
- embedded systems
- design decisions
- pattern classification
- mobile learning
- image classification
- support vector machine svm
- scene understanding
- linear feature extraction
- mobile computing
- mobile applications
- discriminant analysis
- human detection
- energy minimization
- feature selection
- feature space
- feature set
- energy consumption
- background subtraction
- manifold learning
- frequency domain
- mobile environments
- texture analysis
- gabor filters
- context aware
- texture classification
- dimensionality reduction
- local binary pattern
- scene recognition
- feature extractors
- deformable part models