High-Performance Computing
High Performance Mobile Intelligence
- Federated Learning on Mobile Devices
- Distributed Deep Neural Network (DNN) System on Mobile Devices
- Resource-Aware Dynamic DNN Reconfiguration for Mobile Systems
Low Power Mobile Computing Systems
- Comprehensive Power Consumption Analysis for Mobile Systems and Applications
- Performance Retaining Real-time CPU/GPU Scheduling for Mobile Computing Systems
- User Behavior based Mobile Sensor Scheduling
System Security
Intelligence System Robustness, Security
- Algorithm Study for Adversarial Attack and Defense
- DNN Model Robustness Analysis and Interpretation
- Secured Mobile Application Design against Adversarial Attack (e.g., Object Detection, Automatic Speech Recognition, etc.)
Mobile Device Authentication
- Machine Learning based User Finger Operation Behavior Modeling
- Mobile Gesture-based Authentication Schemes
Mobile Display
Organic Ligh Emitting(OLED) Related Display Technologies
- Low Power OLED-based Mobile Display System Design and Optimization
- Circuit Design of OLED Pixel for Fine-grained Dynamic Voltage Scaling
- Sensing & Compensation Circuit Design for OLED Non-uniformity Degradation
Low Power Mobile Display with Human Computer Interaction
- Mobile User Finger-Gaze Behavior Modeling
- HCI-based “FigureShaow” for Low-power Smartphone UI Design
- “Dynamic Display Tone” – A Low-Power OLED UI System Designed for Samsung Galaxy Smartphone Series
Other Projects
Digital Circuit and Memory Related Works
- Circuit Modeling and Analysis for STT-RAM Crossbar Circuits
- Circuit Implementation and Analysis for Neuromorphic Circuits
- Chip Tape-out of a 4K STT-RAM Chip Set — A Joint Project with SMIC and the Chinese Academy of Science Physics Institute