Sony and Raspberry Pi launch powerful new AI camera for developers

Sony Semiconductor Solutions (SSS) and Raspberry Pi Ltd have teamed up to introduce a new product for AI enthusiasts. The Raspberry Pi AI Camera, compatible with all Raspberry Pi single-board computers, is designed to streamline the development of AI solutions by processing visual data directly at the edge. The AI Camera is available today for $70.

This collaboration is a result of SSS’s minority investment in Raspberry Pi Ltd back in April 2023. Since then, the companies have worked together to create an edge AI platform tailored to Raspberry Pi’s developer community. The key feature of the AI Camera is its use of SSS’s IMX500 intelligent vision sensor. This innovative sensor allows AI processing to happen directly on the chip, eliminating the need for additional hardware like GPUs or accelerators, which are traditionally required for handling large-scale visual data.

The AI Camera works with Raspberry Pi’s entire lineup of single-board computers, including the new Raspberry Pi 5. Developers can leverage familiar tools like the libcamera and Picamera2 libraries, making the integration process smoother and more efficient.

Eita Yanagisawa, General Manager of the System Solutions Division at SSS, expressed excitement about this new chapter, saying, “I’m very excited to share SSS edge AI sensing technology with the world’s largest development community as the first fruits of our strategic partnership.”

Eben Upton, CEO of Raspberry Pi Ltd, added, “Together with our longstanding image sensor partner Sony Semiconductor Solutions, we have developed the Raspberry Pi AI Camera, incorporating Sony’s image sensor expertise. We look forward to seeing what our community members are able to achieve using the power of the Raspberry Pi AI Camera.”

The Raspberry Pi AI Camera comes equipped with a 12.3-megapixel IMX500 sensor, capable of capturing high-resolution images at 10 fps and 40 fps depending on the mode. It also includes an RP2040 microcontroller for managing neural network firmware and is pre-loaded with the MobileNetSSD model, making it ready to use out of the box.