RKNN conversion, NPU deployment and camera + AI pipelines.
Export YOLO to ONNX, quantise and convert it to .rknn with RKNN-Toolkit2 on an x86 host, then run it on the board with RKNN-Toolkit-Lite2, using all three RK3588 NPU cores.
Our engineers debug Rockchip boards, BSPs and AI pipelines for product teams. Tell us what you're seeing.
