dump esp-sr v2.0.1

idf_component.yml

@@ -1,4 +1,4 @@
-version: "2.0.0"
+version: "2.0.1"
 description: esp_sr provides basic algorithms for Speech Recognition applications
 url: https://github.com/espressif/esp-sr
 dependencies:

include/esp32/esp_afe_aec.h

@@ -2,9 +2,8 @@
 #ifndef _ESP_AFE_AEC_H_
 #define _ESP_AFE_AEC_H_
 
-
-#include "esp_afe_config.h"
 #include "esp_aec.h"
+#include "esp_afe_config.h"
 
 #include <stdint.h>
 
@@ -13,19 +12,19 @@ extern "C" {
 #endif
 
 typedef struct {
-    aec_handle_t* handle;
+    aec_handle_t *handle;
     aec_mode_t mode;
     afe_pcm_config_t pcm_config;
     int frame_size;
-    int16_t  *data;
-}afe_aec_handle_t;
-
+    int16_t *data;
+} afe_aec_handle_t;
 
 /**
- * @brief Creates an instance to the AEC structure. 
- * 
- * @warning Currently only support 1 microphone channel and 1 playback channe. 
- * If input has multiple microphone channels and playback channels, just the first microphone channel and playback channel will be selected.
+ * @brief Creates an instance to the AEC structure.
+ *
+ * @warning Currently only support 1 microphone channel and 1 playback channe.
+ * If input has multiple microphone channels and playback channels, just the first microphone channel and playback
+ * channel will be selected.
  *
  * The input format, same as afe config:
  * M to represent the microphone channel
@@ -37,25 +36,26 @@ typedef struct {
  *
  * @param input_format     The input format
  * @param filter_length    The length of filter. The larger the filter, the higher the CPU loading.
- *                         Recommended filter_length = 4 for esp32s3 and esp32p4. Recommended filter_length = 2 for esp32c5.
+ *                         Recommended filter_length = 4 for esp32s3 and esp32p4. Recommended filter_length = 2 for
+ * esp32c5.
  * @param type             The type of afe, AFE_TYPE_SR or AFE_TYPE_VC
  * @param mode             The mode of afe, AFE_MODE_LOW_COST or AFE_MODE_HIGH_PERF
  *
  * @return afe_config_t*  The default config of afe
  */
 afe_aec_handle_t *afe_aec_create(const char *input_format, int filter_length, afe_type_t type, afe_mode_t mode);
 
-
 /**
  * @brief Performs echo cancellation a frame, based on the audio sent to the speaker and frame from mic.
- * 
+ *
  * @param inst        The instance of AEC.
- * @param indata      Input audio data, format is define by input_format. Note indata will be modified in function call.
- * @param outdata     Returns near-end signal with echo removed. 
+ * @param indata      Input audio data, format is define by input_format.
+ * @param outdata     Near-end signal with echo removed.  outdata must be 16-bit aligned.
+ *                    please use heap_caps_aligned_calloc(16, n, size, caps) to allocate an aligned chunk of memory
 
  * @return The bytes of outdata.
  */
-size_t afe_aec_process(afe_aec_handle_t *handel, int16_t *indata, int16_t *outdata);
+size_t afe_aec_process(afe_aec_handle_t *handel, const int16_t *indata, int16_t *outdata);
 
 /**
  * @brief Get frame size of AEC (the samples of one frame)
@@ -64,7 +64,6 @@ size_t afe_aec_process(afe_aec_handle_t *handel, int16_t *indata, int16_t *outda
  */
 int afe_aec_get_chunksize(afe_aec_handle_t *handle);
 
-
 /**
  * @brief Free the AEC instance
  *

include/esp32p4/esp_afe_aec.h

@@ -2,9 +2,8 @@
 #ifndef _ESP_AFE_AEC_H_
 #define _ESP_AFE_AEC_H_
 
-
-#include "esp_afe_config.h"
 #include "esp_aec.h"
+#include "esp_afe_config.h"
 
 #include <stdint.h>
 
@@ -13,19 +12,19 @@ extern "C" {
 #endif
 
 typedef struct {
-    aec_handle_t* handle;
+    aec_handle_t *handle;
     aec_mode_t mode;
     afe_pcm_config_t pcm_config;
     int frame_size;
-    int16_t  *data;
-}afe_aec_handle_t;
-
+    int16_t *data;
+} afe_aec_handle_t;
 
 /**
- * @brief Creates an instance to the AEC structure. 
- * 
- * @warning Currently only support 1 microphone channel and 1 playback channe. 
- * If input has multiple microphone channels and playback channels, just the first microphone channel and playback channel will be selected.
+ * @brief Creates an instance to the AEC structure.
+ *
+ * @warning Currently only support 1 microphone channel and 1 playback channe.
+ * If input has multiple microphone channels and playback channels, just the first microphone channel and playback
+ * channel will be selected.
  *
  * The input format, same as afe config:
  * M to represent the microphone channel
@@ -37,25 +36,26 @@ typedef struct {
  *
  * @param input_format     The input format
  * @param filter_length    The length of filter. The larger the filter, the higher the CPU loading.
- *                         Recommended filter_length = 4 for esp32s3 and esp32p4. Recommended filter_length = 2 for esp32c5.
+ *                         Recommended filter_length = 4 for esp32s3 and esp32p4. Recommended filter_length = 2 for
+ * esp32c5.
  * @param type             The type of afe, AFE_TYPE_SR or AFE_TYPE_VC
  * @param mode             The mode of afe, AFE_MODE_LOW_COST or AFE_MODE_HIGH_PERF
  *
  * @return afe_config_t*  The default config of afe
  */
 afe_aec_handle_t *afe_aec_create(const char *input_format, int filter_length, afe_type_t type, afe_mode_t mode);
 
-
 /**
  * @brief Performs echo cancellation a frame, based on the audio sent to the speaker and frame from mic.
- * 
+ *
  * @param inst        The instance of AEC.
- * @param indata      Input audio data, format is define by input_format. Note indata will be modified in function call.
- * @param outdata     Returns near-end signal with echo removed. 
+ * @param indata      Input audio data, format is define by input_format.
+ * @param outdata     Near-end signal with echo removed.  outdata must be 16-bit aligned.
+ *                    please use heap_caps_aligned_calloc(16, n, size, caps) to allocate an aligned chunk of memory
 
  * @return The bytes of outdata.
  */
-size_t afe_aec_process(afe_aec_handle_t *handel, int16_t *indata, int16_t *outdata);
+size_t afe_aec_process(afe_aec_handle_t *handel, const int16_t *indata, int16_t *outdata);
 
 /**
  * @brief Get frame size of AEC (the samples of one frame)
@@ -64,7 +64,6 @@ size_t afe_aec_process(afe_aec_handle_t *handel, int16_t *indata, int16_t *outda
  */
 int afe_aec_get_chunksize(afe_aec_handle_t *handle);
 
-
 /**
  * @brief Free the AEC instance
  *

include/esp32s3/esp_afe_aec.h

@@ -2,9 +2,8 @@
 #ifndef _ESP_AFE_AEC_H_
 #define _ESP_AFE_AEC_H_
 
-
-#include "esp_afe_config.h"
 #include "esp_aec.h"
+#include "esp_afe_config.h"
 
 #include <stdint.h>
 
@@ -13,19 +12,19 @@ extern "C" {
 #endif
 
 typedef struct {
-    aec_handle_t* handle;
+    aec_handle_t *handle;
     aec_mode_t mode;
     afe_pcm_config_t pcm_config;
     int frame_size;
-    int16_t  *data;
-}afe_aec_handle_t;
-
+    int16_t *data;
+} afe_aec_handle_t;
 
 /**
- * @brief Creates an instance to the AEC structure. 
- * 
- * @warning Currently only support 1 microphone channel and 1 playback channe. 
- * If input has multiple microphone channels and playback channels, just the first microphone channel and playback channel will be selected.
+ * @brief Creates an instance to the AEC structure.
+ *
+ * @warning Currently only support 1 microphone channel and 1 playback channe.
+ * If input has multiple microphone channels and playback channels, just the first microphone channel and playback
+ * channel will be selected.
  *
  * The input format, same as afe config:
  * M to represent the microphone channel
@@ -37,25 +36,26 @@ typedef struct {
  *
  * @param input_format     The input format
  * @param filter_length    The length of filter. The larger the filter, the higher the CPU loading.
- *                         Recommended filter_length = 4 for esp32s3 and esp32p4. Recommended filter_length = 2 for esp32c5.
+ *                         Recommended filter_length = 4 for esp32s3 and esp32p4. Recommended filter_length = 2 for
+ * esp32c5.
  * @param type             The type of afe, AFE_TYPE_SR or AFE_TYPE_VC
  * @param mode             The mode of afe, AFE_MODE_LOW_COST or AFE_MODE_HIGH_PERF
  *
  * @return afe_config_t*  The default config of afe
  */
 afe_aec_handle_t *afe_aec_create(const char *input_format, int filter_length, afe_type_t type, afe_mode_t mode);
 
-
 /**
  * @brief Performs echo cancellation a frame, based on the audio sent to the speaker and frame from mic.
- * 
+ *
  * @param inst        The instance of AEC.
- * @param indata      Input audio data, format is define by input_format. Note indata will be modified in function call.
- * @param outdata     Returns near-end signal with echo removed. 
+ * @param indata      Input audio data, format is define by input_format.
+ * @param outdata     Near-end signal with echo removed.  outdata must be 16-bit aligned.
+ *                    please use heap_caps_aligned_calloc(16, n, size, caps) to allocate an aligned chunk of memory
 
  * @return The bytes of outdata.
  */
-size_t afe_aec_process(afe_aec_handle_t *handel, int16_t *indata, int16_t *outdata);
+size_t afe_aec_process(afe_aec_handle_t *handel, const int16_t *indata, int16_t *outdata);
 
 /**
  * @brief Get frame size of AEC (the samples of one frame)
@@ -64,7 +64,6 @@ size_t afe_aec_process(afe_aec_handle_t *handel, int16_t *indata, int16_t *outda
  */
 int afe_aec_get_chunksize(afe_aec_handle_t *handle);
 
-
 /**
  * @brief Free the AEC instance
  *

include/esp32s3/esp_sr_webrtc.h

@@ -0,0 +1,84 @@
+// Copyright 2015-2019 Espressif Systems (Shanghai) PTE LTD
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License
+#ifndef _ESP_WEBRTC_H_
+#define _ESP_WEBRTC_H_
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+#include "esp_agc.h"
+#include "esp_log.h"
+#include "esp_ns.h"
+#include "sr_ringbuf.h"
+#include <stdint.h>
+
+#include "esp_heap_caps.h"
+
+typedef struct {
+    void *ns_handle;
+    void *agc_handle;
+    int frame_size;
+    int sample_rate;
+    int16_t *buff;
+    int16_t *out_data;
+    sr_ringbuf_handle_t rb;
+} webrtc_handle_t;
+
+/**
+ * @brief Creates an instance of webrtc.
+ *
+ * @warning frame_length can supports be 10 ms, 20 ms, 30 ms, 32 ms.
+ *
+ * @param frame_length_ms    The length of the audio processing
+ * @param ns_mode            The mode of NS. -1 means NS is disabled. 0: Mild, 1: Medium, 2: Aggressive
+ * @param agc_mode           The model of AGC
+ * @param agc_gain           The gain of AGC. default is 9
+ * @param agc_target_level   The target level of AGC. default is -3 dbfs
+ * @param sample_rate        The sample rate of the audio.
+ *
+ * @return
+ *         - NULL: Create failed
+ *         - Others: The instance of webrtc
+ */
+webrtc_handle_t *webrtc_create(
+    int frame_length_ms, int ns_mode, agc_mode_t agc_mode, int agc_gain, int agc_target_level, int sample_rate);
+
+/**
+ * @brief Feed samples of an audio stream to the webrtc and get the audio stream after Noise suppression.
+ *
+ * @param handle        The instance of NS.
+ * @param in_data       An array of 16-bit signed audio samples.
+ * @param out_size      The sample size of output data
+ * @param enable_ns     Enable noise suppression
+ * @param enable_agc    Enable automatic gain control
+ *
+ * @return data after noise suppression
+ */
+int16_t *webrtc_process(webrtc_handle_t *handle, int16_t *indata, int *size, bool enable_ns, bool enable_agc);
+
+/**
+ * @brief Free the webrtc instance
+ *
+ * @param handle The instance of webrtc.
+ *
+ * @return None
+ *
+ */
+void webrtc_destroy(webrtc_handle_t *handle);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif //_ESP_NS_H_