A Low-Power Neuromorphic System for Real-Time Visual Activity Recognition.
Deepak KhoslaRyan UhlenbrockYang ChenPublished in: ISVC (2018)
Keyphrases
- activity recognition
- low power
- low cost
- high speed
- real time
- vlsi architecture
- power consumption
- low power consumption
- human activity recognition
- signal processor
- human activities
- action recognition
- single chip
- wearable sensors
- high definition television
- accelerometer data
- activity analysis
- vlsi circuits
- sensory data
- cmos technology
- logic circuits
- visual features
- smart home
- gate array
- power reduction
- image sensor
- pairwise
- context aware
- vlsi implementation
- learning algorithm
- power dissipation
- machine learning