Activation Functions

Functions
arm_cmsis_nn_status	arm_nn_activation_s16 (const int16_t *input, int16_t *output, const int32_t size, const int32_t left_shift, const arm_nn_activation_type type)
	s16 neural network activation function using direct table look-up
void	arm_relu6_s8 (int8_t *data, uint16_t size)
	s8 ReLU6 function
void	arm_relu_q15 (int16_t *data, uint16_t size)
	Q15 RELU function.
void	arm_relu_q7 (int8_t *data, uint16_t size)
	Q7 RELU function.

Description

Perform activation layers, including ReLU (Rectified Linear Unit), sigmoid and tanh

Function Documentation

arm_nn_activation_s16()

arm_cmsis_nn_status arm_nn_activation_s16	(	const int16_t *	input,
		int16_t *	output,
		const int32_t	size,
		const int32_t	left_shift,
		const arm_nn_activation_type	type
	)

s16 neural network activation function using direct table look-up

Parameters

[in]	input	pointer to input data
[out]	output	pointer to output
[in]	size	number of elements
[in]	left_shift	bit-width of the integer part, assumed to be smaller than 3.
[in]	type	type of activation functions

Returns

The function returns ARM_CMSIS_NN_SUCCESS

Supported framework: TensorFlow Lite for Microcontrollers. This activation function must be bit precise congruent with the corresponding TFLM tanh and sigmoid activation functions

arm_relu6_s8()

void arm_relu6_s8	(	int8_t *	data,
		uint16_t	size
	)

s8 ReLU6 function

Parameters

[in,out]	data	pointer to input
[in]	size	number of elements

arm_relu_q15()

void arm_relu_q15	(	int16_t *	data,
		uint16_t	size
	)

Q15 RELU function.

Parameters

[in,out]	data	pointer to input
[in]	size	number of elements

arm_relu_q7()

void arm_relu_q7	(	int8_t *	data,
		uint16_t	size
	)

Q7 RELU function.

Parameters

[in,out]	data	pointer to input
[in]	size	number of elements