I believe in reproducible research and try to open source tools and code out of my works when possible.

Open-source tools and code

• ICLR’24 ArchLock: Locking DNN Transferability at the Architecture Level with a Zero-Cost Binary Predictor  [Code]

• ASPLOS’24 MicroVSA: An Ultra-Lightweight Vector Symbolic Architecture-based Classifier Library for Tiny Microcontrollers

• NeurIPS’23 LinGCN: Structural Linearized Graph Convolutional Network for Homomorphically Encrypted Inference  [Code]

• ICCV’23 VertexSerum: Poisoning Graph Neural Networks for Link Inference [Code]

• ICCV’23 AutoReP: Automatic ReLU Replacement for Fast Private Network Inference  [Code]

• ICML’23 NNSplitter: An Active Defense Solution to DNN Model via Automated Weight Obfuscation [Code]

• ICML’23 Orchestrating Encoding and Sparsity for Fast Homomorphically Encrypted NN Inference [Code]

• DAC’23 Polynomial Arch. Search Framework for Two-party Computation-based Secure NN Deployment  [Code]

• ICCAD’22 ObfuNAS: A Neural Architecture Search-based DNN Obfuscation Approach  [Code] [IEEE/ACM William J. McCalla ICCAD Best Paper Nomination]

• DAC’22 LeHDC: Learning-Based Hyperdimensional Computing Classifier [Code]
  

• tinyML’22 A Brain-Inspired Low-Dimensional Computing Classifier for Inference on Tiny Devices [Code]

• Security’21 An Adversarial Weight Duplication Attack Framework to Crush DNN in Multi-Tenant FPGA  [Code]

• DAC’21 SGX-FPGA: Trusted Execution Environment for CPU-FPGA Heterogeneous Architecture [Code]