Do you have a question about the Dev Board Micro and is the answer not in the manual?
Questions and answers
Subscribe to Our Youtube Channel
Summary of Contents for Coral Dev Board Micro
Page 1
Board Micro This page shows you how to set up the Coral Dev Board Micro, a microcontroller board that can run TensorFlow Lite models with acceleration on the Coral Edge TPU. With its on-board camera and microphone, this board provides a complete system for embedded machine learning (ML) applications.
Page 2
Note: The Dev Board Micro requires a DC power supply that can deliver 5 V at 2 A. Most computers can power the board just fine with USB, and many applications operate at much lower currents. However, the Edge TPU can produce significant power spikes during inferencing, so if you notice that your computer cannot deliver sufficient power via USB, you should connect the board through a powered USB hub.
Page 3
Figure 1. Dev Board Micro hardware interfaces As shown in figure 1, the board includes the following hardware components: Color camera (324 x 324 px) PDM microphone 2 buttons: User programmable button (also toggles power with a 7-second hold) MCU reset button...
Page 4
Because the main processor is a microcontroller (MCU), the Dev Board Micro runs small programs without a general-purpose operating system (there's no Linux OS like on other Coral boards). The platform we've built for Dev Board Micro is based on FreeRTOS and we also offer compatibility with Arduino.
Page 5
To start building apps with FreeRTOS, you need to download the Dev Board Micro source code and install some dependencies, as follows: . Clone the coralmicro repo and all submodules: git clone --recurse-submodules -j8 https://github.com/google-coral/coralmicro This could take a few minutes.
Page 6
6. Try multi-core model cascading A unique feature of the Dev Board Micro is its ability to transition between models running on the M4 core and models running on the M7 core and Edge TPU. This is called "multi-core model cascading" because information cascades between models on each MCU core.
Page 7
Stream images over USB: If you're on Linux, you can see a live stream of camera images with pose overlays by running a client-side Python app: . First install the required packages: python3 -m pip install -r apps/multicore_model_cascade/requirements.txt . Then start the Python app: python3 apps/multicore_model_cascade/multicore_model_cascade.py A new window should appear but it will be black because the camera stream appears only when the pose detection model is running.
Next steps You've now experienced the basic workflow, but there's much more to learn about building apps for the Dev Board Micro. We recommend the following next steps: Connect to the serial console : The serial console is where you can see the standard output from your apps, which is often helpful for debugging.
Page 9
Browse our trained models : Most of these Edge TPU models are compatible with the Dev Board Micro, as indicated by the "Micro" column in each table. However, some models are not compatible because they are too large for the board's...
Page 10
NXP Semiconductors SE Blank RT Family Whereas when the board boots normally, the device appears as: Google Inc. Coral Dev Board Micro STATE_FLASHTOOL_ERROR or No devices detected When flashing the board, if the output includes right before...
Page 11
ls -l /dev/ttyACM0 In the output, the group owner is the 4th column item. For example, "dialout" in this case: crw-rw-rw- 1 root dialout 166, 0 May 26 14:19 /dev/ttyACM0 So you must add your username to this group: sudo usermod -aG dialout $USER Then reboot your computer to apply the group change, and try flashing again.
Need help?
Do you have a question about the Dev Board Micro and is the answer not in the manual?
Questions and answers