TPU-Posenet

Edge TPU Accelerator/Multi-TPU/Multi-Model + Posenet/DeeplabV3/MobileNet-SSD + Python + Sync/Async + LaptopPC/RaspberryPi.
Inspired by google-coral/project-posenet.
This repository was tuned to speed up Google's sample logic to support multi-TPU. And I replaced the complex Gstreamer implementation with the OpenCV implementation.

0. Table of contents

1. Environment
2. Inference behavior
　2-1. Async, TPU x3, USB Camera, Single Person
　2-2. Sync, TPU x1, USB Camera, Single Person
　2-3. Sync, TPU x1, MP4 (30 FPS), Multi Person
　2-4. Async, TPU x3, USB Camera (30 FPS), Multi-Model, Posenet + DeeplabV3 + MobileNet-SSDv2
3. Introduction procedure
　3-1. Common procedures for devices
　3-2-1. Only Linux
　3-2-2. Only RaspberryPi (Stretch or Buster)
4. Usage
5. Reference articles

1. Environment

• Ubuntu or RaspberryPi
  • (Note: Because RaspberryPi3 is a low-speed USB 2.0, multi-TPU operation becomes extremely unstable.)
• OpenCV4.1.1-openvino
• Coral Edge TPU Accelerator (Multi-TPU)
  • Automatically detect the number of multiple TPU accelerators connected to a USB hub to improve performance.
• USB Camera (Playstationeye)
• Picamera v2
• Self-powered USB 3.0 Hub
• Python 3.5.2+

2. Inference behavior

2-1. Async, TPU x3, USB Camera, Single Person

Youtube：https://youtu.be/LBk71RKca1c

2-2. Sync, TPU x1, USB Camera, Single Person

Youtube：https://youtu.be/GuuXzpLXFJo

2-3. Sync, TPU x1, MP4 (30 FPS), Multi Person

Youtube：https://youtu.be/ibPuI12bj2w

2-4. Async, TPU x3, USB Camera (30 FPS), Multi-Model, Posenet + DeeplabV3 + MobileNet-SSDv2

Youtube：https://youtu.be/d946VOE65tU

3. Introduction procedure

3-1. Common procedures for devices

$ sudo apt-get update;sudo apt-get upgrade -y

$ sudo apt-get install -y python3-pip
$ sudo pip3 install pip --upgrade
$ sudo pip3 install numpy

$ wget https://dl.google.com/coral/edgetpu_api/edgetpu_api_latest.tar.gz -O edgetpu_api.tar.gz --trust-server-names
$ tar xzf edgetpu_api.tar.gz
$ sudo edgetpu_api/install.sh

$ git clone https://github.com/PINTO0309/TPU-Posenet.git
$ cd TPU-Posenet.git
$ cd models;./download.sh;cd ..
$ cd media;./download.sh;cd ..

3-2-1. Only Linux

$ wget https://github.com/PINTO0309/OpenVINO-bin/raw/master/Linux/download_2019R2.sh
$ chmod +x download_2019R2.sh
$ ./download_2019R2.sh
$ l_openvino_toolkit_p_2019.2.242/install_openvino_dependencies.sh
$ ./install_GUI.sh
OR
$ ./install.sh

3-2-2. Only RaspberryPi (Stretch or Buster)

### Only Raspbian Buster ############################################################
$ cd /usr/local/lib/python3.7/dist-packages/edgetpu/swig/
$ sudo cp \
_edgetpu_cpp_wrapper.cpython-35m-arm-linux-gnueabihf.so \
_edgetpu_cpp_wrapper.cpython-37m-arm-linux-gnueabihf.so
### Only Raspbian Buster ############################################################

$ cd ~/TPU-Posenet
$ sudo pip3 install imutils
$ sudo raspi-config

$ wget https://github.com/PINTO0309/OpenVINO-bin/raw/master/RaspberryPi/download_2019R2.sh
$ sudo chmod +x download_2019R2.sh
$ ./download_2019R2.sh
$ echo "source /opt/intel/openvino/bin/setupvars.sh" >> ~/.bashrc
$ source ~/.bashrc

4. Usage

usage: pose_camera_multi_tpu.py [-h] [--model MODEL] [--usbcamno USBCAMNO]
                                [--videofile VIDEOFILE] [--vidfps VIDFPS]

optional arguments:
  -h, --help            show this help message and exit
  --model MODEL         Path of the detection model.
  --usbcamno USBCAMNO   USB Camera number.
  --videofile VIDEOFILE
                        Path to input video file. (Default="")
  --vidfps VIDFPS       FPS of Video. (Default=30)

usage: pose_camera_single_tpu.py [-h] [--model MODEL] [--usbcamno USBCAMNO]
                                 [--videofile VIDEOFILE] [--vidfps VIDFPS]

optional arguments:
  -h, --help            show this help message and exit
  --model MODEL         Path of the detection model.
  --usbcamno USBCAMNO   USB Camera number.
  --videofile VIDEOFILE
                        Path to input video file. (Default="")
  --vidfps VIDFPS       FPS of Video. (Default=30)

usage: pose_picam_multi_tpu.py [-h] [--model MODEL] [--videofile VIDEOFILE] [--vidfps VIDFPS]

optional arguments:
  -h, --help            show this help message and exit
  --model MODEL         Path of the detection model.
  --videofile VIDEOFILE
                        Path to input video file. (Default="")
  --vidfps VIDFPS       FPS of Video. (Default=30)

usage: pose_picam_single_tpu.py [-h] [--model MODEL] [--videofile VIDEOFILE] [--vidfps VIDFPS]

optional arguments:
  -h, --help            show this help message and exit
  --model MODEL         Path of the detection model.
  --videofile VIDEOFILE
                        Path to input video file. (Default="")
  --vidfps VIDFPS       FPS of Video. (Default=30)

usage: ssd-deeplab-posenet.py [-h] [--pose_model POSE_MODEL]
                              [--deep_model DEEP_MODEL]
                              [--ssd_model SSD_MODEL] [--usbcamno USBCAMNO]
                              [--videofile VIDEOFILE] [--vidfps VIDFPS]
                              [--camera_width CAMERA_WIDTH]
                              [--camera_height CAMERA_HEIGHT]

optional arguments:
  -h, --help            show this help message and exit
  --pose_model POSE_MODEL
                        Path of the posenet model.
  --deep_model DEEP_MODEL
                        Path of the deeplabv3 model.
  --ssd_model SSD_MODEL
                        Path of the mobilenet-ssd model.
  --usbcamno USBCAMNO   USB Camera number.
  --videofile VIDEOFILE
                        Path to input video file. (Default="")
  --vidfps VIDFPS       FPS of Video. (Default=30)
  --camera_width CAMERA_WIDTH
                        USB Camera resolution (width). (Default=640)
  --camera_height CAMERA_HEIGHT
                        USB Camera resolution (height). (Default=480)

5. Reference articles

1. Edge TPU USB Accelerator analysis - I/O data transfer - Qiita - iwatake2222

2. [150 FPS ++] Connect three Coral Edge TPU accelerators and infer in parallel processing to get ultra-fast object detection inference performance ーTo the extreme of useless high performanceー - Qiita - PINTO

3. [150 FPS ++] Connect three Coral Edge TPU accelerators and infer in parallel processing to get ultra-fast Posenet inference performance ーTo the extreme of useless high performanceー - Qiita - PINTO

4. Raspberry Pi Camera Module