Yet another development board with RP2040 SoC, ArduCam has announced its Arducam Pico4ML dedicated to machine learning applications. The silicon-based RP2040 SoC being a cost-effective solution to support edge computing with low-power specifications. We also saw Arduino’s RP2040 Nano Connect Board which aims at combining Arduino’s libraries with Raspberry Pi’s optimized SoC architecture. It also featured some sensors which could be helpful for building a wide range of use cases but, nothing specific to machine learning or Edge AI.
CNX Software’s generic post on machine learning with RP2040 has unveiled some interesting points regarding the Arducam Pico4ML. The low-power RP2040 chip could be slow while running the machine learning inference as observed in its previous use case for the person detection using TensorFlow Lite. On the contrary, the company may have designed Arduam Pico4ML to solve this issue as the selling point of the board lies in its ML use cases. Hence, it would be worth exploring the board to analyze for coming to a conclusion.
Discussing further, the Arducam board integrates the TensorFlow Lite Micro framework, which would ease the developers to train models for specific use cases. Also, the community support for TensorFlow Lite Micro can speed up the application building. Additionally, the board comes with some instances for beginners to use TensorFlow Lite Micro on the Arducam Pico4ML.
ArduCam says “It is a good platform for endpoint AI, or more specifically TinyML. Hence, enabling users to run machine learning (ML) models to do sensor-based analysis. This includes voice and image recognition and accelerometer-based gesture recognition.”
Arducam Pico4ML also features:
- Wake word detection
- Magic wand
- Person detection
You can also observe the on-board LCD display for the person detection use case in the featured youtube video for the Arducam Pico4ML board. The use case features real-time person detection using its inbuilt sensors.
The pricing information is not available as the board is not released at this point in time. So, you can join the waitlist or follow the Twitter updates of the Arducam Pico4ML board. For more information visit the official product page. Images and technical specifications have also been taken from the product page.
Saumitra Jagdale is a Backend Developer, Freelance Technical Author, Global AI Ambassador (SwissCognitive), Open-source Contributor in Python projects, Leader of Tensorflow Community India and Passionate AI/ML Enthusiast