Download PDFOpen PDF in browser

Support Vector Machine Hardware Accelerator for Tongue Color Diagnosis

EasyChair Preprint no. 6903

9 pagesDate: October 20, 2021

Abstract

Tongue body features such as color is used in Traditional Chinese Medicine (TCM) practices to diagnose a patient's state of health. However, the diagnosis of one patient's health condition varies between practitioners. This gives rise to the need of a standard method such as using Support Vector Machine (SVM) to identify the tongue body color. Typically, SVM classifiers are implemented in software where the classification performance is very dependent on the architecture of the general-purpose CPU. Since classification of tongue images is a recurring event, the design of a hardware accelerator is proposed in this project. The purpose of designing a hardware accelerator is to boost the classifier performance, execution time and latency so that it meets real-time constraints. Architectural optimization methods, such as loop unrolling, memory array partitioning and pipelining of the SVM classification algorithm are utilized and the final hardware architecture is synthesized to Xilinx Virtex-7 FPGA. To further optimize the resource utilization, 18-bits IEEE-754 floating-point representations for the floating-point units are used. The proposed SVM hardware demonstrates rougly 140x speed up with only 1% of classification accuracy loss when compared to the software implementation in MATLAB.

Keyphrases: FPGA, Hardware Accelerator, hardware-software comparison, Support Vector Machine, tongue colour diagnosis

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@Booklet{EasyChair:6903,
  author = {Amanda Thiah Su Lin and Mohd Shahrizal Rusli and Nur Diyana Kamarudin and Ab Al-Hadi Ab Rahman and Usman Ullah Sheikh and Michael Tan Loong Peng and Shahidatul Sadiah and Mohd Ibrahim Shapiai},
  title = {Support Vector Machine Hardware Accelerator for Tongue Color Diagnosis},
  howpublished = {EasyChair Preprint no. 6903},

  year = {EasyChair, 2021}}
Download PDFOpen PDF in browser