Coral
  • Products
    • Product gallery
    • Prototyping
    • Production
    • Sensors
    • Partner products
    • Technology
  • Industries
    • Our industries
    • Smart cities
    • Manufacturing
    • Automotive
    • Healthcare
    • Agriculture
  • Examples
    • Code examples
    • Partner examples
    • Project tutorials
  • Docs & Tools
    • Documentation
    • Models
    • Software
  • Support
  • About
    • About Coral
    • News
    • Partnerships
    • Sales
    • Legal
  • search close
  • Hardware
  • Dev Board
    • Get started
    • Connect to a camera
    • Connect to the I/O pins
    • Connect to the serial console
    • Update or flash the board
    • Datasheet
  • Dev Board Mini
    • Get started
    • Connect to a camera
    • Connect to the I/O pins
    • Connect to the serial console
    • Update or flash the board
    • Datasheet
  • USB Accelerator
    • Get started
    • Datasheet
  • System-on-Module
    • Get started
    • Datasheet
  • M.2 & Mini PCIe Accelerator
    • Get started
    • M.2 Datasheet
    • Mini PCIe Datasheet
    • M.2 with Dual Edge TPU Datasheet
  • Accelerator Module
    • Datasheet
  • Environmental Sensor
    • Get started
    • Datasheet
    • coral.enviro.board
    • coral.cloudiot.core
  • Camera
    • Datasheet
  • Tools
  • Mendel Development Tool
  • Edge TPU Compiler
  • Running a model
  • Inferencing overview
  • Run inference with Python
  • Run inference with C++
  • Run multiple models with multiple Edge TPUs
  • Pipeline a model with multiple Edge TPUs
  • Creating a model
  • TensorFlow models overview
  • Retrain a classification model in Google Colab
  • Retrain a classification model in Docker
  • Retrain an object detection model in Docker
  • Retrain a classification model on-device with weight imprinting
  • Retrain a classification model on-device with backpropagation
  • API Reference
  • PyCoral API (Python)
    • Overview
    • pycoral.utils
    • pycoral.adapters
    • pycoral.pipeline
    • pycoral.learn.backprop
    • pycoral.learn.imprinting
  • Libcoral API (C++)
    • Overview
    • Inferencing APIs
    • Pipelining APIs
    • Transfer learning APIs
  • Libedgetpu API (C++)
  • Edge TPU Python API (deprecated)
    • Overview
    • edgetpu.basic.basic_engine
    • edgetpu.basic.edgetpu_utils
    • edgetpu.classification.engine
    • edgetpu.detection.engine
    • edgetpu.learn.backprop.softmax_regression
    • edgetpu.learn.imprinting.engine
    • edgetpu.utils.dataset_utils
    • edgetpu.utils.image_processing
  • Application notes
  • Build Coral for your platform
  • Manage the PCIe module temperature
  • Appendix
  • Performance benchmarks
  • Frequently asked questions

Documentation

Learn how to build AI products with Coral devices

Set up your device

  • Dev Board
  • Dev Board Mini
  • USB Accelerator
  • PCIe/M.2 Accelerator

Perform an inference

  • Inferencing overview
  • Try image classification
  • Try object detection
  • More examples

Create your own model

  • Model compatibility
  • Retrain a classification model
  • Retrain an object detection model

Downloads

  • Software
  • Pre-compiled models

Source code

  • Mendel OS
  • Software libraries

Support

  • FAQ
  • Get help
  • About Coral
  • News
  • Google AI principles
  • Terms of service
  • Privacy policy
  • Documentation
  • Examples
  • Support
  • Contact sales
  • Legal

Copyright 2020 Google LLC. All rights reserved.

Stay up-to-date

Sign up to receive our newsletter

Subscribe
You are subscribed! Thanks!
Close Icon Created with Sketch.