14.6 A 0.62mW ultra-low-power convolutional-neural-network face-recognition processor and a CIS integrated with always-on haar-like face detector.
Kyeongryeol BongSungpill ChoiChanghyeon KimSanghoon KangYouchang KimHoi-Jun YooPublished in: ISSCC (2017)
Keyphrases
- face detection
- convolutional neural network
- face detector
- face recognition
- human faces
- haar like features
- detection accuracy
- detection rate
- facial images
- power consumption
- ultra low power
- face databases
- detection method
- low power
- face images
- object detection
- adaboost algorithm
- facial features
- facial region
- skin color
- high speed
- computer vision
- recognition rate
- single chip
- face tracking
- feature extraction
- pose variations
- principal component analysis
- visual tracking
- frontal face
- real time
- facial expressions
- detection algorithm
- sparse representation