speech_recognition/esp-sr
1
0
Fork 0
mirror of https://github.com/espressif/esp-sr.git synced 2025-09-15 15:28:44 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

enlarge MASE filter_strength range & lower processed freq range

Browse Source
This commit is contained in:
Zhao Yang 2020-05-13 20:06:59 +08:00
parent 87bb8b5cf3
commit 4d0a595652
3 changed files with 30 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
acoustic_algorithm/include/esp_mase.h
View File

@ -48,13 +48,13 @@ typedef void* mase_handle_t;
* *
* @param operating_mode '0' for normal mode and '1' for wake-up enhanced mode. * @param operating_mode '0' for normal mode and '1' for wake-up enhanced mode.
* *
* @param filter_strength Strengh of the mic-array speech enhancement, must be 0, 1, 2 or 3. * @param filter_strength Strengh of the mic-array speech enhancement, must be 0 to 6.
* *
* @return * @return
* - NULL: Create failed * - NULL: Create failed
* - Others: An instance of MASE * - Others: An instance of MASE
*/ */
mase_handle_t mase_create(int sample_rate, int frame_size, int array_type, float mic_distance, int operating_mode, int filter_strength); mase_handle_t mase_create(int fs, int frame_size, int array_type, float mic_distance, int operating_mode, int filter_strength);
/** /**
* @brief Performs mic array processing for one frame. * @brief Performs mic array processing for one frame.
@ -67,9 +67,6 @@ mase_handle_t mase_create(int sample_rate, int frame_size, int array_type, float
* *
* @return None * @return None
* *
* @note Input is a multi-channel signal while the output is single-channel.
* For a 16-ms multi-channel input frame, the i-th point in the c-th channel should be indexed (i + c * 256).
*
*/ */
void mase_process(mase_handle_t st, int16_t *in, int16_t *dsp_out); void mase_process(mase_handle_t st, int16_t *in, int16_t *dsp_out);

BIN
acoustic_algorithm/libesp_audio_processor.a
View File

Binary file not shown.

28
main/audio_process.c
View File

@ -9,6 +9,7 @@
#include "esp_aec.h" #include "esp_aec.h"
#include "esp_agc.h" #include "esp_agc.h"
#include "esp_vad.h" #include "esp_vad.h"
#include "esp_mase.h"
#include "audio_test_file.h" #include "audio_test_file.h"
void NSTask(void *arg) void NSTask(void *arg)
@ -105,6 +106,31 @@ void VADTask(void *arg)
vTaskDelete(NULL); vTaskDelete(NULL);
} }
void MASETask(void *arg)
{
mase_handle_t mase_st = mase_create(MASE_SAMPLE_RATE, MASE_FRAME_SIZE, THREE_MIC_CIRCLE, MASE_MIC_DISTANCE, WAKE_UP_ENHANCEMENT_MODE, 6);
int chunks = 0;
int nch = 3;
int audio_chunksize = MASE_FRAME_SIZE * MASE_SAMPLE_RATE * nch / 1000;
int16_t *mase_in = malloc(audio_chunksize * sizeof(int16_t));
int16_t *mase_out = malloc(audio_chunksize / nch * sizeof(int16_t));
while (1) {
if ((chunks + 1) * audio_chunksize * sizeof(int16_t) <= sizeof(audio_test_file)) {
memcpy(mase_in, audio_test_file + chunks * audio_chunksize * sizeof(int16_t), audio_chunksize * sizeof(int16_t));
memset(mase_out, 0, audio_chunksize / nch * sizeof(int16_t));
} else {
break;
}
mase_process(mase_st, mase_in, mase_out);
chunks++;
}
mase_destory(mase_st);
free(mase_in);
free(mase_out);
printf("MASE test successfully\n\n");
printf("TEST3 FINISHED\n\n");
vTaskDelete(NULL);
}
void audio_process_test() void audio_process_test()
{ {
@ -115,4 +141,6 @@ void audio_process_test()
xTaskCreatePinnedToCore(&AECTask, "acoustic_echo_cancellation", 3 * 1024, NULL, 5, NULL, 0); xTaskCreatePinnedToCore(&AECTask, "acoustic_echo_cancellation", 3 * 1024, NULL, 5, NULL, 0);
vTaskDelay(1000 / portTICK_PERIOD_MS); vTaskDelay(1000 / portTICK_PERIOD_MS);
xTaskCreatePinnedToCore(&VADTask, "voice_activity_detection", 3 * 1024, NULL, 5, NULL, 0); xTaskCreatePinnedToCore(&VADTask, "voice_activity_detection", 3 * 1024, NULL, 5, NULL, 0);
vTaskDelay(1000 / portTICK_PERIOD_MS);
xTaskCreatePinnedToCore(&MASETask, "mic_array_speech_enhancement", 3 * 1024, NULL, 5, NULL, 1);
} }