科学研究

[TOSEM]PatchCensor: Patch Robustness Certification for Transformers via Exhaustive Testing

[ICCD 2022] Nesting Forward Automatic Differentiation in DNN for Memory-Efficient Framework Design

[MobiCom 2022] MobiDepth: Real-Time Depth Estimation Using On-Device Dual Cameras

[TMC] LEAP: TrustZone Based Developer-Friendly TEE for Intelligent Mobile Apps

​

[TMC] FLASH: Heterogeneity-Aware Federated Learning at Scale

[SenSys 2022] Hyperion: A Generic and Distributed Mobile Offloading Framework on OpenCL

[IEEE TMC] S-UbiTap: Leveraging Acoustic Dispersion for Ubiquitous and Scalable Touch Interface on Solid Surf...

[MICRO 2022] ANT: Exploiting Adaptive Numerical Data Type for Low-bit Deep Neural Network Quantization

[MobiSys 2022] FedBalancer: Data and Pace Control for Efficient Federated Learning on Heterogeneous Clients

[MobiSys 2022] Breaking the Memory Wall for On-Device Learning

[MobiSys 2022] CoDL: Efficient CPU-GPU Co-execution for Deep Learning Inference on Mobile Devices

[TMC 2022] MVPose: Realtime Multi-Person Pose Estimation using Motion Vector on Mobile Devices

[JSAC 2022] WheelLoc: Practical and Accurate Localization for Wheeled Mobile Target via Integrated Sensing an...

[ICSE 2022] ReMoS: Reducing Defect Inheritance in Transfer Learning via Relevant Model Slicing

[ICLR 2022] SQuant: On-the-Fly Data-Free Quantization via Diagonal Hessian Approximation