Edge Impulse’s YOLO-Pro has been shortlisted for the Electronics Excellence Awards under the Software category.
YOLO-Pro is a new object-detection architecture built for real-time Edge deployment, delivering YOLO-class accuracy with compute, latency, and power requirements that are optimised for the key types of Edge hardware on the market – along with clear, commercial-friendly licensing.
The original YOLO line has been instrumental in advancing computer vision, offering fast and accurate detection for research and Cloud-based workflows. But these generic object detection models were not designed for embedded systems: they are designed to beat academic benchmarks that are not representative of real world tasks, they are optimised only for GPU-class acceleration, and they often carry restrictive or ambiguous licenses (AGPL, custom proprietary terms, or unclear commercial allowances).
YOLO-Pro was made to address the needs of developer and enterprises doing computer vision on the Edge. It is the first YOLO architecture built specifically for industrial, embedded, and IoT deployment. Engineered by Edge Impulse from the ground up, YOLO-Pro is available in multiple sizes – from ultra-small (~682K parameters) to high-accuracy large variants – and optimised to run efficiently on MCUs, CPUs, NPUs, GPUs, and Edge appliances. Developers can choose between SiLU architecture for maximum accuracy or a lightweight ReLU-based variant for ultra-efficient inference on constrained devices.
For more information, visit: https://www.edgeimpulse.com/blog/introducing-yolo-pro-object-detection-optimized-for-the-edge/
The Electronics Excellence Awards ceremony will take place at embedded world on 11th March at 4.30pm at the Hall 5 Forum.
You’re all welcome to come and see which product will win!
