XCORE-VISION combines a compact hardware platform with an integrated camera and a growing collection of open-source software designed to make embedded computer vision easy to explore and quick to prototype.

Powered by the XMOS XCORE.AI multicore microcontroller, the platform delivers deterministic, real-time performance through hardware-level parallel processing,without requiring an operating system or cloud connectivity.

The result is a platform that sits comfortably between simple MCU camera modules and heavyweight Linux systems, making it ideal for developers who need responsive, low-power AI at the edge.

Explore fully documented app notes running MobileNetV2 and Yolov8 directly on the board for applications such as:

Object classification
Face identification
Logo detection
Multi-modal AI combining vision and audio
Custom image recognition models

Everything runs locally on the board, providing predictable latency, privacy, and complete control over the application.