Low Power Inference for On-Device Visual Recognition with A Quantization-Friendly Solution

Authors:
  • CFChen Feng
    Chen Feng
  • TSTao Sheng
    Tao Sheng
  • ZLZhiyu Liang
    Zhiyu Liang
  • SZShaojie Zhuo
    Shaojie Zhuo
  • XZXiaopeng Zhang
    Xiaopeng Zhang
  • LSLiang Shen
    Liang Shen
  • MAMatthew Ardi
    Matthew Ardi
  • ABAlexander C. Berg
    Alexander C. Berg
  • YCYiran Chen
    Yiran Chen
  • BCBo Chen
    Bo Chen
  • KGKent Gauen
    Kent Gauen
  • YLYung-Hsiang Lu
    Yung-Hsiang Lu

Abstract
The IEEE Low-Power Image Recognition Challenge (LPIRC) is an annual competition started in 2015 that encourages joint hardware and software solutions for computer vision systems with low latency and power. Track 1 of the competition in 2018 focused on the innovation of software solutions with fixed inference engine and hardware. This decision allows participants to submit models online and not worry about building and bringing custom hardware on-site, which attracted a historically large number of submissions. Among the diverse solutions, the winning solution proposed a quantization-friendly framework for MobileNets that achieves an accuracy of 72.67% on the holdout dataset with an average latency of 27ms on a single CPU core of Google Pixel2 phone, which is superior to the best real-time MobileNet models at the time.