Merge branch 'feat/support_two_wn' into 'master'

Feat/support two wn

See merge request speech-recognition-framework/esp-sr!50

CHANGELOG.md

@@ -4,6 +4,11 @@
 - Available storage is less than the remaining flash space on IDF v5.0.   
 If you can not map model partition successfully, please check the left free storage by `spi_flash_mmap_get_free_pages(ESP_PARTITION_MMAP_DATA)` or update IDF to v5.1 or later.
 
+## 1.4.0
+- Add ci tests to check the APIs of wakenet, multinet and AFE work fine
+- Support to load and run two wakenet9 models at the same time in AFE
+- Reduce the latency of multinet6
+
 ## 1.3.4
 - Fix the bug of multinet5q8 whrn adding new speech commands
 

idf_component.yml

@@ -1,4 +1,4 @@
-version: "1.3.4"
+version: "1.4.0"
 description: esp_sr provides basic algorithms for Speech Recognition applications
 url: https://github.com/espressif/esp-sr
 dependencies:
@@ -7,7 +7,8 @@ dependencies:
 files:
   exclude:
     - ".github"
-    - "docs"
+    - "docs/**/*"
+    - "test_apps/**/*"
 targets:
   - esp32
   - esp32s2

include/esp32s3/esp_afe_config.h

@@ -64,7 +64,8 @@ typedef struct {
     bool voice_communication_agc_init;      // AGC swich for voice communication
     int voice_communication_agc_gain;       // AGC gain(dB) for voice communication
     vad_mode_t vad_mode;                    // The value can be: VAD_MODE_0, VAD_MODE_1, VAD_MODE_2, VAD_MODE_3, VAD_MODE_4
-    char *wakenet_model_name;               // The model name of wakenet
+    char *wakenet_model_name;               // The model name of wakenet 1
+    char *wakenet_model_name_2;             // The model name of wakenet 2 if has wakenet 2
     det_mode_t wakenet_mode;
     afe_sr_mode_t afe_mode;
     int afe_perferred_core;
@@ -89,6 +90,7 @@ typedef struct {
     .voice_communication_agc_gain = 15, \
     .vad_mode = VAD_MODE_3, \
     .wakenet_model_name = NULL, \
+    .wakenet_model_name_2 = NULL, \
     .wakenet_mode = DET_MODE_90, \
     .afe_mode = SR_MODE_HIGH_PERF, \
     .afe_perferred_core = 0, \
@@ -114,6 +116,7 @@ typedef struct {
     .voice_communication_agc_gain = 15, \
     .vad_mode = VAD_MODE_3, \
     .wakenet_model_name = NULL, \
+    .wakenet_model_name_2 = NULL, \
     .wakenet_mode = DET_MODE_2CH_90, \
     .afe_mode = SR_MODE_LOW_COST, \
     .afe_perferred_core = 0, \

include/esp32s3/esp_afe_sr_iface.h

@@ -27,6 +27,7 @@ typedef struct afe_fetch_result_t
     int data_size;                          // the size of data. The unit is byte.
     wakenet_state_t wakeup_state;           // the value is wakenet_state_t
     int wake_word_index;                    // if the wake word is detected. It will store the wake word index which start from 1.
+    int wakenet_model_index;                // if there are multiple wakenets, this value identifies which model be wakes up. Index start from 1.
     afe_vad_state_t vad_state;              // the value is afe_vad_state_t
     int trigger_channel_id;                 // the channel index of output
     int wake_word_length;                   // the length of wake word. It's unit is the number of samples.
@@ -111,7 +112,7 @@ typedef int (*esp_afe_sr_iface_op_reset_buffer_t)(esp_afe_sr_data_t *afe);
 
 /**
  * @brief Initial wakenet and wake words coefficient, or reset wakenet and wake words coefficient 
- *        when wakenet has been initialized.  
+ *        when wakenet has been initialized. It's only support wakenet 1 now.
  *
  * @param afe                The AFE_SR object to query
  * @param wakenet_word       The wakenet word, should be DEFAULT_WAKE_WORD or EXTRA_WAKE_WORD