Luckfox Forums

i just bought luckfox rv1106 yesterday, I just want to know if luckfox rv1106 supports using tensorflow lite or lite micro?

It seems like currently only Mobilenetv2 and YoloV5 well supported on this board, which aren't bad models.

For faster inference mobilenet, for more accurate inference Yolov5.

grayfacenospace wrote: ↑2024-01-23 18:31 It seems like currently only Mobilenetv2 and YoloV5 well supported on this board, which aren't bad models.

For faster inference mobilenet, for more accurate inference Yolov5.

how to use Mobbilenetv2 on luckfox?

Tensorflow Lite should be supported using the RKNN Toolkit2
https://github.com/rockchip-linux/rknn-toolkit2

CPP examples are here https://github.com/rockchip-linux/rknpu ... 106_RV1103

tortuga wrote: ↑2024-01-26 9:48 Tensorflow Lite should be supported using the RKNN Toolkit2
https://github.com/rockchip-linux/rknn-toolkit2

According to the latest repo only mobilenet and yolov5 are compatible right now. Those are already rather a bit complicated to use, I doubt tflite is easy to get working

https://github.com/airockchip/rknn_model_zoo

"RKNN Model Zoo is developed based on the RKNPU SDK toolchain and provides deployment examples for current mainstream algorithms. Include the process of exporting the RKNN model and using Python API and CAPI to infer the RKNN model.

Support RK3562, RK3566, RK3568, RK3588 platforms. (RV1103, RV1106 platforms support mobilenet, yolov5)