Overview / Usage

• Environment
  • LattePanda Alpha (Intel 7th Core m3-7y30) or LaptopPC (Intel 8th Core i7-8750H)
  • Ubuntu 16.04 x86_64
  • RaspberryPi3
  • Raspbian Stretch armv7l
  • OpenVINO toolkit 2018 R5 (2018.5.445)
  • Python 3.5
  • C++
  • OpenCV 4.0.1-openvino
  • Tensorflow v1.11.0 or Tensorflow-GPU v1.11.0 (pip install)
  • YoloV3 (MS-COCO)
  • tiny-YoloV3 (MS-COCO)
  • USB Camera (PlaystationEye) / Movie file (mp4)
  • Intel Neural Compute Stick v1 / v2
    　
• Environment construction procedure
  https://github.com/PINTO0309/OpenVINO-YoloV3#environment-construction-procedure
  　
• How to check the graph structure of a ".pb" file
  https://github.com/PINTO0309/OpenVINO-YoloV3#how-to-check-the-graph-structure-of-a-pb-file-part4
  　
• Issue
  How to offload OpenVINO non-compliant layer to Tensorflow (undefined symbol: _ZN15InferenceEngine10TensorDescC1Ev)
  OpenVINO failing on YoloV3's YoloRegion, only one working on FP16, all working on FP32
  　
• Future tasks
  • Work that makes compatibility with Multiple models

Methodology / Approach

OpenVINO

Technologies Used

OpenVINO+YoloV3/tiny-YoloV3+NCS/NCS2

Repository

https://github.com/PINTO0309/OpenVINO-YoloV3

Other links

• [13 FPS] NCS2 x 4 + Full size YoloV3 performance has been tripled
• [24 FPS] Boost RaspberryPi3 with four Neural Compute Stick 2 (NCS2) MobileNet-SSD / YoloV3 [48 FPS for Core i7]

Collaborators