Edge Impulse Announces Official Espressif ESP32 Support, Releases Open Source ESP-EYE Firmware

Initial firmware offers a 3-4x speedup over stock TensorFlow Lite for Microcontrollers, with optimizations to follow.

Edge machine learning specialist Edge Impulse has announced a major new string to its bow: official Edge Impulse Studio support for the Espressif ESP32 family, focusing primarily on the ESP-EYE artificially intelligent camera board.

"We have selected the ESP-EYE as a reference for Edge Impulse Studio integration since it is equipped with an OV2640 camera, microphone, 8MB of PSRAM, and 4MB of flash memory, allowing for the development of voice, vision, and other sensor data-enabled machine learning applications," expains Edge Impulse's Dmitry Maslov. "Having said that, both pre-built firmware and a source code for it can be useful for connecting other ESP32-based boards to the Edge Impulse ecosystem."

Edge Impulse Studio now supports ESP32 devices, courtesy of an open source ESP-EYE firmware release. (πŸ“Ή: Edge Impulse)

The support for the board includes the release of an open source Edge Impulse-specific firmware for the ESP-EYE, which includes data collection and inference capabilities for the on-board microphone and camera β€” but, Maslov claims, can be easily expanded to other ESP32-based boards including the ESP-CAM.

"ESP32 represents a new milestone for Edge Impulse not only because of its immense popularity in the community, but also because it is the first major non-Arm processor board supported," says Maslov. "It has brought a set of unique challenges when working on firmware for ESP32 β€” the CMSIS-NN and CMSIS-DSP optimizations could have not been used for the Tensilica Xtensa LX6 microprocessor."

Instead, Edge Impulse on ESP32 chips uses Espressif's ESP-NN neural network acceleration functionality, which offers a 3-4x performance boost over stock TensorFlow Lite for Microcontrollers. "That," Maslov points out, "allows running our keyword spotting project real-time with high accuracy."

The initial firmware release includes support for Omnivision OV2640, OV3660, and OV5640 image sensors, the ESP-EYE's on-board I2S microphone, an STMicroelectronics LIS3DHTR accelerometer connected to the I2C bus, and "any analog sensor" connected to the ESP32's A0 pin.

So far there's no word on plans to add extra sensor support, but Maslov has confirmed ESP-DSP will be added to the next firmware release to accelerate spectral analysis, MFCC, and MFE digital signal processing blocks in Edge Impulse Studio projects.

More information is available on the Edge Impulse blog, while those eager to try the new firmware out can find it on the company's GitHub repository under a permissive open source license.

Gareth Halfacree
Freelance journalist, technical author, hacker, tinkerer, erstwhile sysadmin. For hire: freelance@halfacree.co.uk.
Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles