• Coral offers a platform of hardware components (Tensor Processing Unit: TPU), software tools, and pre-compiled models for building devices with local AI.
• Coral USB Accelerator is a USB device . It provides an Edge TPU as a coprocessor for your computer.
• How is it powered?
  • C USB A is powered by 5 V from the USB interface. It is powered through on-board power management chip (PMIC).
• How to set up C USB A?
  • On the computer where you'll connect the USB Accelerator

(Ref: https://pollen-robotics.github.io/reachy-docs/docs/appendices/install-from-scratch-raspberry/)

Instructions Stepwise:

• Download the Edge TPU runtime
  • Commands: (for unpacking & installing Edge TPU runtime)
    • echo "deb https://packages.cloud.google.com/apt coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list
    • curl https://packages.cloud.google.com/apt/doc/apt-key.gpg
    • sudo apt-get update
    • sudo apt-get install libedgetpu1-std
  • Decisions:
    • Do not enable the maximum operating frequency if you’re unsure of the performance needed for your application. Heat issues are cons.
• Connect the USB Accelerator to the computer with the USB 3.0 cable
• Download the TensorFlow Lite library :
  • (Ref https://coral.ai/docs/accelerator/get-started/#2-install-the-tensorflow-lite-library)
  • Install the tflite_runtime library (only the TFlite interpreter is needed, not the full TF library. Only a part of it, the TFLite interpreter)
    • Commands:
      • pip3 install https://dl.google.com/coral/python/tflite_runtime-2.1.0.post1-cp37-cp37m-linux_armv7l.whl (For Python 2.7, or on Rasp 10 OS on Rasp pi)
      • Open label_image.py file & Edit 2 lines of the code
        • import tflite_runtime.interpreter as tflite
        • interpreter = tflite.Interpreter(model_path=args.model_file)
      • Run label_image.py
• You are set up to run TensorFlow Lite models.
• (For further details, scroll to the bottom here: https://www.tensorflow.org/lite/guide/python)
• Perform image classification with our example code and model
  • Commands:
    • mkdir coral && cd coral
    • git clone https://github.com/google-coral/tflite.git
    • cd tflite/python/examples/classification
    • bash install_requirements.sh
    • python3 classify_image.py
    • ‘’’(more details on python code are here: https://github.com/google-coral/tflite/blob/master/python/examples/classification/classify_image.py)’’’
    • (You should results: Ara macao (Scarlet Macaw): 0.76562)