lib/libtf: Update libtf regression to support 2D input.

updated tf_regregression to take ulab array as input

updated tf_regregression to take ulab array as input

updated libtf_regression function name

included ulab ndarray

updated libtf header file
This commit is contained in:
tejalbarnwal 2023-06-25 05:13:50 +05:30
parent cc434bbee0
commit 2d8f3f94a1
2 changed files with 15 additions and 15 deletions

View File

@ -1,5 +1,5 @@
/* This file is part of the OpenMV project.
* Copyright (c) 2013-2021 Ibrahim Abdelkader <iabdalkader@openmv.io> & Kwabena W. Agyeman <kwagyeman@openmv.io>
* Copyright (c) 2013-2023 Ibrahim Abdelkader <iabdalkader@openmv.io> & Kwabena W. Agyeman <kwagyeman@openmv.io>
* This work is licensed under the MIT license, see the file LICENSE for details.
*/
@ -72,7 +72,8 @@ int libtf_generate_micro_features(const int16_t *input, // Audio samples
int8_t *output, // Slice data
size_t *num_samples_read); // Number of samples used
int libtf_regression_1Dinput_1Doutput(const unsigned char *model_data, uint8_t* tensor_arena, libtf_parameters_t* params, float* input_data, float* output_data);
// runs regression on 2D/ 1D input(provided as array) and return 1D output
int libtf_regression(const unsigned char *model_data, uint8_t* tensor_arena, libtf_parameters_t* params, float* input_data, float* output_data);
#ifdef __cplusplus
}

View File

@ -13,12 +13,14 @@
#include "py/obj.h"
#include "py/objlist.h"
#include "py/objtuple.h"
#include "py/objarray.h"
#include "py/binary.h"
#include "py_helper.h"
#include "imlib_config.h"
#include "ulab/code/ulab.h"
#include "ulab/code/ndarray.h"
#ifdef IMLIB_ENABLE_TF
#include "py_image.h"
#include "ff_wrapper.h"
@ -266,30 +268,27 @@ STATIC mp_obj_t py_tf_regression(uint n_args, const mp_obj_t *args, mp_map_t *kw
// read model
py_tf_model_obj_t *arg_model = py_tf_load_alloc(args[0]);
size_t input_size = (&arg_model->params)->input_width;
// read input(2D or 1D) and output size(1D)
size_t input_size_width = (&arg_model->params)->input_width;
size_t input_size_height = (&arg_model->params)->input_height;
size_t output_size = (&arg_model->params)->output_channels;
// read input
mp_obj_array_t *arg_input_array = args[1];
ndarray_obj_t *arg_input_array = args[1];
// check for the input size
if (input_size != arg_input_array->len) {
if ((input_size_width * input_size_height) != arg_input_array->len) {
mp_raise_msg(&mp_type_ValueError, MP_ERROR_TEXT("Input array size is not same as model input size!"));
}
float input_array[input_size];
for (size_t i=0; i<input_size; i++) {
input_array[i] = (float) mp_obj_float_get(
mp_binary_get_val_array(arg_input_array->typecode, arg_input_array->items, i)
);
}
float *input_array = (float *)(arg_input_array->array);
uint8_t *tensor_arena = fb_alloc(arg_model->params.tensor_arena_size, FB_ALLOC_PREFER_SPEED | FB_ALLOC_CACHE_ALIGN);
float output_data[output_size];
// predict the output using tflite model
if (libtf_regression_1Dinput_1Doutput(arg_model->model_data,
if (libtf_regression(arg_model->model_data,
tensor_arena, &arg_model->params, input_array, output_data) != 0){
mp_raise_msg(&mp_type_OSError, MP_ERROR_TEXT("Coundnt execute the model to predict the output"));
}