Welcome to the Wisconsin Embedded Systems and Computing (WISEST) Lab at the University of Wisconsin–Madison. We are engaged in research on design, architecture, and optimization of hardware and software of embedded computing systems. We are particularly interested in machine learning at the edge, energy-quality scalability, mobile and embedded security, and Internet-of-Things.
Machine Learning at the Edge
Machine learning on edge devices provides the benefits of shorter latency, less network dependency, and enhanced privacy, but extremely energy-efficient hardware is required due to the severe energy constraint.
- Hardware optimization for low-power neural networks
Approximate computing is a promising approach to the energy-efficient computing of error-resilient applications where the quality of computation results can be traded for energy efficiency, such as machine learning and signal processing.
- SAADI(-EC): Accuracy-scalable divider unit
- SECO: Accuracy-scalable exponential function unit
- AxSerBus: Accuracy-scalable serial bus
Security- and privacy-sensitive applications and services on mobile and embedded computing platforms need low-power, low-cost security measures.
- Velody: User authentication using vibration ▶
- VoltKey: Context-based device authentication using power line noise
- MicPrint: Spoof-resistant device authentication using microphone sensor fingerprint
- CamPUF: Device authentication using image sensor fingerprint
Smart wearable devices
Real-time monitoring and control in industrial and agricultural environments facilitate the optimal operation of the system and improve productivity and environmental sustainability in various domains.
- Human activity recognition on wearable computer
- Real-time monitoring of dairy cattle body temperature