feat(mn): Add multinet5q8 Chinese model

include/esp32s3/esp_mn_iface.h

@@ -7,6 +7,10 @@
 #define ESP_MN_MAX_PHRASE_LEN 63
 #define ESP_MN_MIN_PHRASE_LEN 2
 
+#define ESP_MN_PREFIX "mn"
+#define ESP_MN_ENGLISH "en"
+#define ESP_MN_CHINESE "cn"
+
 typedef enum {
 	ESP_MN_STATE_DETECTING = 0,     // detecting
 	ESP_MN_STATE_DETECTED = 1,      // detected
@@ -84,6 +88,14 @@ typedef int (*esp_mn_iface_op_set_det_threshold_t)(model_iface_data_t *model, fl
  */
 typedef int (*esp_mn_iface_op_get_samp_rate_t)(model_iface_data_t *model);
 
+/**
+ * @brief Get the language of model
+ *
+ * @param model       The language name 
+ * @return Language name string defined in esp_mn_models.h, eg: ESP_MN_CHINESE, ESP_MN_ENGLISH
+ */
+typedef char * (*esp_mn_iface_op_get_language_t)(model_iface_data_t *model);
+
 /**
  * @brief Feed samples of an audio stream to the speech recognition model and detect if there is a speech command found.
  *
@@ -133,6 +145,7 @@ typedef struct {
     esp_mn_iface_op_get_samp_chunksize_t get_samp_chunksize;
     esp_mn_iface_op_get_samp_chunknum_t get_samp_chunknum;
     esp_mn_iface_op_set_det_threshold_t set_det_threshold;
+    esp_mn_iface_op_get_language_t get_language;
     esp_mn_iface_op_detect_t detect; 
     esp_mn_iface_op_destroy_t destroy;
     esp_mn_iface_op_get_results_t get_results;

include/esp32s3/esp_mn_models.h

@@ -5,11 +5,6 @@
 //a specific phrase or word.
 
 
-
-#define ESP_MN_PREFIX "mn"
-#define ESP_MN_ENGLISH "en"
-#define ESP_MN_CHINESE "cn"
-
 /**
  * @brief Get the multinet handle from model name
  *

model/movemodel.py

@@ -64,6 +64,8 @@ elif "CONFIG_SR_MN_CN_MULTINET4_5_SINGLE_RECOGNITION_QUANT8" in MN_STRING:
     multinet_model.append('mn4q8_cn')
 elif "CONFIG_SR_MN_CN_MULTINET4_5_SINGLE_RECOGNITION" in MN_STRING and len(multinet_model) < 2:
     multinet_model.append('mn4_cn')
+elif "CONFIG_SR_MN_CN_MULTINET5_RECOGNITION_QUANT8" in MN_STRING and len(multinet_model) < 2:
+    multinet_model.append('mn5q8_cn')
 
 if "CONFIG_SR_MN_EN_MULTINET5_SINGLE_RECOGNITION_QUANT8" in MN_STRING and len(multinet_model) < 2:
     multinet_model.append('mn5q8_en')

model/multinet_model/mn5q8_cn/_MODEL_INFO_

@@ -0,0 +1,2 @@
+# (neural network type)_(model data version)_(lable1_detection windown length_threshold for 90%_threshold for 95%)_(lable2 ...)_...
+MN5Q8_v2_chinese_8_0.9_0.90

src/esp_process_sdkconfig.c

@@ -54,8 +54,7 @@ void check_chip_config(void)
 
 char *get_id_name_cn(int i)
 {
-#if defined CONFIG_USE_MULTINET
+#if defined CONFIG_SR_MN_CN_MULTINET2_SINGLE_RECOGNITION || defined CONFIG_SR_MN_CN_MULTINET4_5_SINGLE_RECOGNITION || defined CONFIG_SR_MN_CN_MULTINET4_5_SINGLE_RECOGNITION_QUANT8 || defined CONFIG_SR_MN_CN_MULTINET5_RECOGNITION_QUANT8
-#if defined CONFIG_SR_MN_CN_MULTINET2_SINGLE_RECOGNITION || defined CONFIG_SR_MN_CN_MULTINET4_5_SINGLE_RECOGNITION || defined CONFIG_SR_MN_CN_MULTINET4_5_SINGLE_RECOGNITION_QUANT8 || defined CONFIG_SR_MN_CN_MULTINET5_SINGLE_RECOGNITION_QUANT8
     if (i == 0)
         return CONFIG_CN_SPEECH_COMMAND_ID0;
     else if (i == 1)
@@ -461,9 +460,6 @@ char *get_id_name_cn(int i)
 #else
     return NULL;
 #endif
-#else
-    return NULL;
-#endif
 }
 
 char *get_id_name_en(int i)
@@ -879,20 +875,34 @@ char *get_id_name_en(int i)
 esp_mn_error_t* esp_mn_commands_update_from_sdkconfig(const esp_mn_iface_t *multinet,  model_iface_data_t *model_data)
 {
     esp_mn_commands_alloc();
+    printf("esp_mn_commands_update_from_sdkconfig\n");
     int total_phrase_num = 0;
+    int language_id = 1; // 0: Chinese, 1:English
+#ifdef CONFIG_SR_MN_CN_MULTINET2_SINGLE_RECOGNITION
+    language_id = 1;
+#else
+    if (strcmp(ESP_MN_CHINESE, multinet->get_language(model_data)) == 0) {
+        language_id = 1;
+    } else if (strcmp(ESP_MN_ENGLISH, multinet->get_language(model_data)) == 0) {
+        language_id = 2;
+    } else {
+        ESP_LOGE(TAG, "Invalid language");
+        return NULL;
+    }
+#endif
 
     for (int i = 0; i < ESP_MN_MAX_PHRASE_NUM; i++) {
         char *command_str = NULL;
         int command_str_len = 0;
 
-#if defined CONFIG_SR_MN_CN_MULTINET2_SINGLE_RECOGNITION || defined CONFIG_SR_MN_CN_MULTINET4_SINGLE_RECOGNITION || defined CONFIG_SR_MN_CN_MULTINET4_5_SINGLE_RECOGNITION || defined CONFIG_SR_MN_CN_MULTINET4_5_SINGLE_RECOGNITION_QUANT8 || defined CONFIG_SR_MN_CN_MULTINET5_SINGLE_RECOGNITION_QUANT8
+        if (language_id == 1) {
-        command_str = get_id_name_cn(i);
+            command_str = get_id_name_cn(i);
-#elif defined CONFIG_SR_MN_EN_MULTINET5_SINGLE_RECOGNITION || defined CONFIG_SR_MN_EN_MULTINET5_SINGLE_RECOGNITION_QUANT8
+        } else if (language_id == 2) {
-        command_str = get_id_name_en(i);
+            command_str = get_id_name_en(i);
-#else
+        } else {
-        ESP_LOGE(TAG, "Incorrect language");
+            ESP_LOGE(TAG, "Invalid language");
-#endif
+        }
-        
+
         if (command_str == NULL) continue;
         command_str_len = strlen(command_str);
         if (command_str_len <= 0) continue;