Edge AI Add-on API 2.2.0
Loading...
Searching...
No Matches
Flatten and sliding by columns window

Data Structures

struct  nrf_dsp_window_flatsliding_s
 

Typedefs

typedef struct nrf_dsp_window_flatsliding_s nrf_dsp_window_flatsliding_t
 

Functions

int8_t nrf_dsp_window_flatsliding_init (nrf_dsp_window_flatsliding_t *p_ctx, void *p_window, const uint16_t max_samples_num, const size_t uniq_features_num, const uint16_t window_shift)
 Initialize flatten by colums sliding window context type.
 
uint16_t nrf_dsp_window_flatsliding_feed_f32 (nrf_dsp_window_flatsliding_t *p_ctx, const flt32_t *p_input, uint16_t num)
 Feeds f32 samples to the floating-point flatten by colums window, with ordering by feature columns, e.g p_input = [x,y,z,x1,y1,z1] -> p_window = [x,x1,.., y,y1,.., z,z1,..].
 
uint16_t nrf_dsp_window_flatsliding_feed_i8 (nrf_dsp_window_flatsliding_t *p_ctx, const int8_t *p_input, uint16_t num)
 Feeds INT8 samples to the INT8 flatten by colums window, with ordering by feature columns, e.g p_input = [x,y,z,x1,y1,z1] -> p_window = [x,x1,.., y,y1,.., z,z1,..].
 
uint16_t nrf_dsp_window_flatsliding_feed_i16 (nrf_dsp_window_flatsliding_t *p_ctx, const int16_t *p_input, uint16_t num)
 Feeds INT16 samples to the INT16 flatten by colums window, with ordering by feature columns, e.g p_input = [x,y,z,x1,y1,z1] -> p_window = [x,x1,.., y,y1,.., z,z1,..].
 
uint16_t nrf_dsp_window_flatsliding_feed_bymask_f32 (nrf_dsp_window_flatsliding_t *p_ctx, const flt32_t *p_input, uint16_t num, const uint8_t *p_usage_mask)
 Feeds f32 samples to the floating-point flatten by colums window, with ordering by feature columns, e.g p_input = [x,y,z,x1,y1,z1] -> p_window = [x,x1,.., y,y1,.., z,z1,..], usage of inputs column is determined by p_input_usage_mask where each bit represents each unique feature.
 
uint16_t nrf_dsp_window_flatsliding_feed_bymask_i8 (nrf_dsp_window_flatsliding_t *p_ctx, const int8_t *p_input, uint16_t num, const uint8_t *p_usage_mask)
 Feeds INT8 samples to the floating-point flatten by colums window, with ordering by feature columns, e.g p_input = [x,y,z,x1,y1,z1] -> p_window = [x,x1,.., y,y1,.., z,z1,..], usage of inputs column is determined by p_input_usage_mask where each bit represents each unique feature.
 
uint16_t nrf_dsp_window_flatsliding_feed_bymask_i16 (nrf_dsp_window_flatsliding_t *p_ctx, const int16_t *p_input, uint16_t num, const uint8_t *p_usage_mask)
 Feeds INT16 samples to the floating-point flatten by colums window, with ordering by feature columns, e.g p_input = [x,y,z,x1,y1,z1] -> p_window = [x,x1,.., y,y1,.., z,z1,..], usage of inputs column is determined by p_input_usage_mask where each bit represents each unique feature.
 

Detailed Description