I would like to try to deploy model beside from the computer visions ones shown from the demo in application hub.
I would like to know how the npu is being declared or used in the demo codes. The demo codes don't seems to directly interact with the npu. Are there any APIs or function to utilize the on-chip npu? Or is the npu automatically managed?
已解决! 转到解答。
Hi @TomC818
You're correct that in the MCXN947 demos, the NPU is not typically accessed through direct API calls in the application code. Instead, the NPU is automatically invoked by the TensorFlow Lite for Microcontrollers runtime.
You can refer to the dm-multiple-person-detection-on-mcxn947 demo.
BR
Harry
Hi @TomC818
You're correct that in the MCXN947 demos, the NPU is not typically accessed through direct API calls in the application code. Instead, the NPU is automatically invoked by the TensorFlow Lite for Microcontrollers runtime.
You can refer to the dm-multiple-person-detection-on-mcxn947 demo.
BR
Harry