devttl/xiaozhi-esp32
1
0
Fork 0
forked from xiaozhi/xiaozhi-esp32
Code Pull Requests Activity

Detect wake word model from index.json (#1211)

Browse Source 
* detect wake word model from index.json

* update wait time before entering wifi configure mode
This commit is contained in:
Xiaoxia
2025-09-17 08:31:51 +08:00
committed by GitHub
parent a1e1f73886
commit f418c16b2c
20 changed files with 75 additions and 134 deletions
Download Patch File Download Diff File Expand all files Collapse all files

63
main/audio/audio_service.cc
View File

@@ -8,12 +8,11 @@
#include "processors/no_audio_processor.h"
#endif
#if CONFIG_USE_AFE_WAKE_WORD
#if CONFIG_IDF_TARGET_ESP32S3 || CONFIG_IDF_TARGET_ESP32P4
#include "wake_words/afe_wake_word.h"
#elif CONFIG_USE_ESP_WAKE_WORD
#include "wake_words/esp_wake_word.h"
#elif CONFIG_USE_CUSTOM_WAKE_WORD
#include "wake_words/custom_wake_word.h"
#else
#include "wake_words/esp_wake_word.h"
#endif
#define TAG "AudioService"
@@ -50,16 +49,6 @@ void AudioService::Initialize(AudioCodec* codec) {
audio_processor_ = std::make_unique<NoAudioProcessor>();
#endif
#if CONFIG_USE_AFE_WAKE_WORD
wake_word_ = std::make_unique<AfeWakeWord>();
#elif CONFIG_USE_ESP_WAKE_WORD
wake_word_ = std::make_unique<EspWakeWord>();
#elif CONFIG_USE_CUSTOM_WAKE_WORD
wake_word_ = std::make_unique<CustomWakeWord>();
#else
wake_word_ = nullptr;
#endif
audio_processor_->OnOutput([this](std::vector<int16_t>&& data) {
PushTaskToEncodeQueue(kAudioTaskTypeEncodeToSendQueue, std::move(data));
});
@@ -71,14 +60,6 @@ void AudioService::Initialize(AudioCodec* codec) {
}
});
if (wake_word_) {
wake_word_->OnWakeWordDetected([this](const std::string& wake_word) {
if (callbacks_.on_wake_word_detected) {
callbacks_.on_wake_word_detected(wake_word);
}
});
}
esp_timer_create_args_t audio_power_timer_args = {
.callback = [](void* arg) {
AudioService* audio_service = (AudioService*)arg;
@@ -100,11 +81,11 @@ void AudioService::Start() {
#if CONFIG_USE_AUDIO_PROCESSOR
/* Start the audio input task */
xTaskCreate([](void* arg) {
xTaskCreatePinnedToCore([](void* arg) {
AudioService* audio_service = (AudioService*)arg;
audio_service->AudioInputTask();
vTaskDelete(NULL);
}, "audio_input", 2048 * 3, this, 8, &audio_input_task_handle_);
}, "audio_input", 2048 * 3, this, 8, &audio_input_task_handle_, 0);
/* Start the audio output task */
xTaskCreate([](void* arg) {
@@ -670,4 +651,36 @@ void AudioService::CheckAndUpdateAudioPowerState() {
void AudioService::SetModelsList(srmodel_list_t* models_list) {
models_list_ = models_list;
}
#if CONFIG_IDF_TARGET_ESP32S3 || CONFIG_IDF_TARGET_ESP32P4
if (esp_srmodel_filter(models_list_, ESP_MN_PREFIX, NULL) != nullptr) {
wake_word_ = std::make_unique<CustomWakeWord>();
} else if (esp_srmodel_filter(models_list_, ESP_WN_PREFIX, NULL) != nullptr) {
wake_word_ = std::make_unique<AfeWakeWord>();
} else {
wake_word_ = nullptr;
}
#else
if (esp_srmodel_filter(models_list_, ESP_WN_PREFIX, NULL) != nullptr) {
wake_word_ = std::make_unique<EspWakeWord>();
} else {
wake_word_ = nullptr;
}
#endif
if (wake_word_) {
wake_word_->OnWakeWordDetected([this](const std::string& wake_word) {
if (callbacks_.on_wake_word_detected) {
callbacks_.on_wake_word_detected(wake_word);
}
});
}
}
bool AudioService::IsAfeWakeWord() {
#if CONFIG_IDF_TARGET_ESP32S3 || CONFIG_IDF_TARGET_ESP32P4
return wake_word_ != nullptr && dynamic_cast<AfeWakeWord*>(wake_word_.get()) != nullptr;
#else
return false;
#endif
}

1
main/audio/audio_service.h
View File

@@ -94,6 +94,7 @@ public:
bool IsIdle();
bool IsWakeWordRunning() const { return xEventGroupGetBits(event_group_) & AS_EVENT_WAKE_WORD_RUNNING; }
bool IsAudioProcessorRunning() const { return xEventGroupGetBits(event_group_) & AS_EVENT_AUDIO_PROCESSOR_RUNNING; }
bool IsAfeWakeWord();
void EnableWakeWordDetection(bool enable);
void EnableVoiceProcessing(bool enable);

2
main/audio/wake_words/custom_wake_word.cc
View File

@@ -91,8 +91,10 @@ bool CustomWakeWord::Initialize(AudioCodec* codec, srmodel_list_t* models_list)
if (models_list == nullptr) {
language_ = "cn";
models_ = esp_srmodel_init("model");
#if CONFIG_CUSTOM_WAKE_WORD
threshold_ = CONFIG_CUSTOM_WAKE_WORD_THRESHOLD / 100.0f;
commands_.push_back({CONFIG_CUSTOM_WAKE_WORD, CONFIG_CUSTOM_WAKE_WORD_DISPLAY, "wake"});
#endif
} else {
models_ = models_list;
ParseWakenetModelConfig();