FPGA-based Domain Specific Accelerators

2020/02 - Present

• An FPGA-based flexible detection system for gutter oil is proposed for the first time based onmulti-dimensional features and machine learning.
• An optimized iris recognition architecture is proposed and deployed to FPGA to realize real-time iris recognition;
• An FPGA-based highly-integrated wearable edge device thatlocally recognizes hand gestures, rather than uploading data to the cloud as in previous studies.
• Publications:
  1. “Gutter oil detection for food safety based on multi-feature machine learning and implementation on FPGA with approximate multipliers”, PeerJ Comput. Sci., 2021, 7: e774.
  2. “Low-power Iris Recognition System Implementation on FPGA with Approximate Multiplier,” J. Comput, 2021, 32: 115-127.
  3. “Wearable on-device deep learning system for hand gesture recognition based on FPGA accelerator,” Math. Biosci. Eng., 2021, 18(1): 132-153.