Neuton inference API

Run neural network inference for Neuton models. More...

Functions
nrf_edgeai_err_t	nrf_edgeai_run_inference_neuton_q8 (nrf_edgeai_t *p_edgeai)
nrf_edgeai_err_t	nrf_edgeai_run_inference_neuton_q16 (nrf_edgeai_t *p_edgeai)
nrf_edgeai_err_t	nrf_edgeai_run_inference_neuton_f32 (nrf_edgeai_t *p_edgeai)

Detailed Description

Run neural network inference for Neuton models.

These functions perform a forward pass (inference) of the neural network using Neuton models with different numeric representations for activations, and neuron buffers:

• 8-bit quantized
• 16-bit quantized
• 32-bit floating point

The input source (raw input or extracted features) is selected based on model meta information, and computes the output by propagating features through all neurons.

Parameters

[in,out]

p_edgeai

Pointer to the Edge AI solution context (nrf_edgeai_t).

Returns

Operation status code nrf_edgeai_err_t