Rdzleo
e95d0c414e
实现 ESP32-S3 上单摄像头人脸追踪的核心代码骨架，替代 Grove Vision AI V2
模块，通过 UART 发送人脸坐标驱动 RP2040 控制的眼球/YAW 舵机。
## 规划文档（docs/phase-01-face-tracking/）
- GOAL.md Phase 目标与 5 大成功标准
- RESEARCH.md esp-dl v3.2/3.3 + human_face_detect 0.4.1 技术调研
- PLAN.md 15 个原子任务的执行计划（T01-T15）
- PLAN_CHECK.md 计划审查报告（PASS_WITH_NOTES）
- PROGRESS.md 执行进度追踪（批次 1-3 已完成）
## 批次 1：依赖与开关（T01-T03）
- main/idf_component.yml
新增 esp-dl ~3.3.0 + human_face_detect 0.4.1（仅 S3/P4）
esp-sr 从 ~2.2.0 升级到 ~2.3.1，解决 esp-dsp 1.6/1.7 版本冲突
- main/Kconfig.projbuild
新增 CONFIG_XIAOZHI_ENABLE_FACE_TRACKING 开关（默认 y，depends on S3）
新增 CONFIG_XIAOZHI_FACE_TRACKING_FPS_CHOICE（5/10/15）
- main/boards/common/esp32_camera.{h,cc}
新增 ProbeFrameCapture() 最小 V4L2 DQBUF/QBUF 探针（T01）
- main/application.cc
Start() 末尾调用 probe 验证摄像头硬件链路
## 批次 2：人脸检测核心（T04-T06）
- main/boards/common/esp32_camera.{h,cc}
新增 FrameRef 结构体 + CaptureForDetection/ReleaseDetectionFrame
双超时 mutex 策略：face_tracker 10ms timeout 跳帧，Capture() RAII guard
- main/face_tracker.{h,cc}（新建）
Core 0 / 优先级 2 / 栈 8KB 独立任务
集成 esp-dl HumanFaceDetect 推理
坐标归一化 cx*224/W-112，匹配 RP2040 pixel_centre=112
多人脸遍历挑 score 最高，避免多脸时眼球摇摆
三重保护：Kconfig depends on S3 + 源文件 #if 守卫 + CMake 条件排除
- main/CMakeLists.txt
非 S3 目标从 SOURCES 移除 face_tracker.cc
## 批次 3：UART 协议扩展（T07）
- main/uart_component.{h,cc}
新增 uart_send_face(x,y) 发送 face:x,y\r\n 协议
extern "C" 链接名配合 face_tracker 的弱符号声明
全局 TX mutex 保护所有 UART 写入，防并发帧交织
uart_send_string 同步加锁保持一致性
## 编译验证
idf.py build 通过，固件 2.51MB / 剩余 1.46MB (36% free)
当前 face_tracker 未被 application 激活（留到 T11），
UART/摄像头现有功能零影响。
## 未完成（下次继续）
- T01 硬件 probe 实机验证
- T08-T10 RP2040 端 parse_face + facetrack 双数据源改造
- T11-T15 application 接入 + 端到端联调 + 性能调优 + 最终验收
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
933 lines
34 KiB
C++
933 lines
34 KiB
C++
#include "application.h"
|
||
#include "board.h"
|
||
#include "display.h"
|
||
#include "system_info.h"
|
||
#include "audio_codec.h"
|
||
#include "mqtt_protocol.h"
|
||
#include "websocket_protocol.h"
|
||
#include "assets/lang_config.h"
|
||
#include "mcp_server.h"
|
||
#include "assets.h"
|
||
#include "settings.h"
|
||
|
||
#include <cstring>
|
||
#include <esp_log.h>
|
||
#include <cJSON.h>
|
||
#include <driver/gpio.h>
|
||
#include <arpa/inet.h>
|
||
#include <font_awesome.h>
|
||
#include <uart_component.h>
|
||
|
||
// [T01] 临时 probe：验证 OV3660 + esp_video 底层采集链路（Phase 01）
|
||
// 仅在非 ESP32（原版）目标上可用——esp32_camera 组件本身也是这个守卫
|
||
#ifndef CONFIG_IDF_TARGET_ESP32
|
||
#include <esp_timer.h>
|
||
#include "boards/common/esp32_camera.h"
|
||
#endif
|
||
|
||
#define TAG "Application"
|
||
|
||
|
||
// ESP32 设备状态字符串，用于日志输出
|
||
// 其中 "idle"、"listening"、"speaking" 会通过 UART 发送给 RP2040 驱动舵机动画
|
||
static const char* const STATE_STRINGS[] = {
|
||
"unknown",
|
||
"starting",
|
||
"configuring",
|
||
"idle", // → RP2040 执行 state_sleep（闭眼、耳朵收起）
|
||
"connecting",
|
||
"listening", // → RP2040 执行 state_listen（好奇姿态、眼球随机看）
|
||
"speaking", // → RP2040 执行 state_speaking（嘴巴开合、眼球随机看、身体摆动）
|
||
"upgrading",
|
||
"activating",
|
||
"audio_testing",
|
||
"fatal_error",
|
||
"invalid_state"
|
||
};
|
||
|
||
Application::Application() {
|
||
event_group_ = xEventGroupCreate();
|
||
|
||
#if CONFIG_USE_DEVICE_AEC && CONFIG_USE_SERVER_AEC
|
||
#error "CONFIG_USE_DEVICE_AEC and CONFIG_USE_SERVER_AEC cannot be enabled at the same time"
|
||
#elif CONFIG_USE_DEVICE_AEC
|
||
aec_mode_ = kAecOnDeviceSide;
|
||
#elif CONFIG_USE_SERVER_AEC
|
||
aec_mode_ = kAecOnServerSide;
|
||
#else
|
||
aec_mode_ = kAecOff;
|
||
#endif
|
||
|
||
esp_timer_create_args_t clock_timer_args = {
|
||
.callback = [](void* arg) {
|
||
Application* app = (Application*)arg;
|
||
xEventGroupSetBits(app->event_group_, MAIN_EVENT_CLOCK_TICK);
|
||
},
|
||
.arg = this,
|
||
.dispatch_method = ESP_TIMER_TASK,
|
||
.name = "clock_timer",
|
||
.skip_unhandled_events = true
|
||
};
|
||
esp_timer_create(&clock_timer_args, &clock_timer_handle_);
|
||
}
|
||
|
||
Application::~Application() {
|
||
if (clock_timer_handle_ != nullptr) {
|
||
esp_timer_stop(clock_timer_handle_);
|
||
esp_timer_delete(clock_timer_handle_);
|
||
}
|
||
vEventGroupDelete(event_group_);
|
||
}
|
||
|
||
void Application::CheckAssetsVersion() {
|
||
auto& board = Board::GetInstance();
|
||
auto display = board.GetDisplay();
|
||
auto& assets = Assets::GetInstance();
|
||
|
||
if (!assets.partition_valid()) {
|
||
ESP_LOGW(TAG, "Assets partition is disabled for board %s", BOARD_NAME);
|
||
return;
|
||
}
|
||
|
||
Settings settings("assets", true);
|
||
// Check if there is a new assets need to be downloaded
|
||
std::string download_url = settings.GetString("download_url");
|
||
|
||
if (!download_url.empty()) {
|
||
settings.EraseKey("download_url");
|
||
|
||
char message[256];
|
||
snprintf(message, sizeof(message), Lang::Strings::FOUND_NEW_ASSETS, download_url.c_str());
|
||
Alert(Lang::Strings::LOADING_ASSETS, message, "cloud_arrow_down", Lang::Sounds::OGG_UPGRADE);
|
||
|
||
// Wait for the audio service to be idle for 3 seconds
|
||
vTaskDelay(pdMS_TO_TICKS(3000));
|
||
SetDeviceState(kDeviceStateUpgrading);
|
||
board.SetPowerSaveMode(false);
|
||
display->SetChatMessage("system", Lang::Strings::PLEASE_WAIT);
|
||
|
||
bool success = assets.Download(download_url, [display](int progress, size_t speed) -> void {
|
||
std::thread([display, progress, speed]() {
|
||
char buffer[32];
|
||
snprintf(buffer, sizeof(buffer), "%d%% %uKB/s", progress, speed / 1024);
|
||
display->SetChatMessage("system", buffer);
|
||
}).detach();
|
||
});
|
||
|
||
board.SetPowerSaveMode(true);
|
||
vTaskDelay(pdMS_TO_TICKS(1000));
|
||
|
||
if (!success) {
|
||
Alert(Lang::Strings::ERROR, Lang::Strings::DOWNLOAD_ASSETS_FAILED, "circle_xmark", Lang::Sounds::OGG_EXCLAMATION);
|
||
vTaskDelay(pdMS_TO_TICKS(2000));
|
||
return;
|
||
}
|
||
}
|
||
|
||
// Apply assets
|
||
assets.Apply();
|
||
display->SetChatMessage("system", "");
|
||
display->SetEmotion("microchip_ai");
|
||
uart_send_string("configuring");
|
||
}
|
||
|
||
void Application::CheckNewVersion(Ota& ota) {
|
||
const int MAX_RETRY = 10;
|
||
int retry_count = 0;
|
||
int retry_delay = 10; // 初始重试延迟为10秒
|
||
|
||
auto& board = Board::GetInstance();
|
||
while (true) {
|
||
SetDeviceState(kDeviceStateActivating);
|
||
auto display = board.GetDisplay();
|
||
display->SetStatus(Lang::Strings::CHECKING_NEW_VERSION);
|
||
|
||
esp_err_t err = ota.CheckVersion();
|
||
if (err != ESP_OK) {
|
||
retry_count++;
|
||
if (retry_count >= MAX_RETRY) {
|
||
ESP_LOGE(TAG, "Too many retries, exit version check");
|
||
return;
|
||
}
|
||
|
||
char error_message[128];
|
||
snprintf(error_message, sizeof(error_message), "code=%d, url=%s", err, ota.GetCheckVersionUrl().c_str());
|
||
char buffer[256];
|
||
snprintf(buffer, sizeof(buffer), Lang::Strings::CHECK_NEW_VERSION_FAILED, retry_delay, error_message);
|
||
Alert(Lang::Strings::ERROR, buffer, "cloud_slash", Lang::Sounds::OGG_EXCLAMATION);
|
||
|
||
ESP_LOGW(TAG, "Check new version failed, retry in %d seconds (%d/%d)", retry_delay, retry_count, MAX_RETRY);
|
||
for (int i = 0; i < retry_delay; i++) {
|
||
vTaskDelay(pdMS_TO_TICKS(1000));
|
||
if (device_state_ == kDeviceStateIdle) {
|
||
break;
|
||
}
|
||
}
|
||
retry_delay *= 2; // 每次重试后延迟时间翻倍
|
||
continue;
|
||
}
|
||
retry_count = 0;
|
||
retry_delay = 10; // 重置重试延迟时间
|
||
|
||
if (ota.HasNewVersion()) {
|
||
if (UpgradeFirmware(ota)) {
|
||
return; // This line will never be reached after reboot
|
||
}
|
||
// If upgrade failed, continue to normal operation (don't break, just fall through)
|
||
}
|
||
|
||
// No new version, mark the current version as valid
|
||
ota.MarkCurrentVersionValid();
|
||
if (!ota.HasActivationCode() && !ota.HasActivationChallenge()) {
|
||
xEventGroupSetBits(event_group_, MAIN_EVENT_CHECK_NEW_VERSION_DONE);
|
||
// Exit the loop if done checking new version
|
||
break;
|
||
}
|
||
|
||
display->SetStatus(Lang::Strings::ACTIVATION);
|
||
// Activation code is shown to the user and waiting for the user to input
|
||
if (ota.HasActivationCode()) {
|
||
ShowActivationCode(ota.GetActivationCode(), ota.GetActivationMessage());
|
||
}
|
||
|
||
// This will block the loop until the activation is done or timeout
|
||
for (int i = 0; i < 10; ++i) {
|
||
ESP_LOGI(TAG, "Activating... %d/%d", i + 1, 10);
|
||
esp_err_t err = ota.Activate();
|
||
if (err == ESP_OK) {
|
||
xEventGroupSetBits(event_group_, MAIN_EVENT_CHECK_NEW_VERSION_DONE);
|
||
break;
|
||
} else if (err == ESP_ERR_TIMEOUT) {
|
||
vTaskDelay(pdMS_TO_TICKS(3000));
|
||
} else {
|
||
vTaskDelay(pdMS_TO_TICKS(10000));
|
||
}
|
||
if (device_state_ == kDeviceStateIdle) {
|
||
break;
|
||
}
|
||
}
|
||
}
|
||
}
|
||
|
||
void Application::ShowActivationCode(const std::string& code, const std::string& message) {
|
||
struct digit_sound {
|
||
char digit;
|
||
const std::string_view& sound;
|
||
};
|
||
static const std::array<digit_sound, 10> digit_sounds{{
|
||
digit_sound{'0', Lang::Sounds::OGG_0},
|
||
digit_sound{'1', Lang::Sounds::OGG_1},
|
||
digit_sound{'2', Lang::Sounds::OGG_2},
|
||
digit_sound{'3', Lang::Sounds::OGG_3},
|
||
digit_sound{'4', Lang::Sounds::OGG_4},
|
||
digit_sound{'5', Lang::Sounds::OGG_5},
|
||
digit_sound{'6', Lang::Sounds::OGG_6},
|
||
digit_sound{'7', Lang::Sounds::OGG_7},
|
||
digit_sound{'8', Lang::Sounds::OGG_8},
|
||
digit_sound{'9', Lang::Sounds::OGG_9}
|
||
}};
|
||
|
||
// This sentence uses 9KB of SRAM, so we need to wait for it to finish
|
||
Alert(Lang::Strings::ACTIVATION, message.c_str(), "link", Lang::Sounds::OGG_ACTIVATION);
|
||
|
||
for (const auto& digit : code) {
|
||
auto it = std::find_if(digit_sounds.begin(), digit_sounds.end(),
|
||
[digit](const digit_sound& ds) { return ds.digit == digit; });
|
||
if (it != digit_sounds.end()) {
|
||
audio_service_.PlaySound(it->sound);
|
||
}
|
||
}
|
||
}
|
||
|
||
void Application::Alert(const char* status, const char* message, const char* emotion, const std::string_view& sound) {
|
||
ESP_LOGW(TAG, "Alert [%s] %s: %s", emotion, status, message);
|
||
auto display = Board::GetInstance().GetDisplay();
|
||
display->SetStatus(status);
|
||
display->SetEmotion(emotion);
|
||
// Alert 时将情绪字符串（如 "happy"、"neutral"）发送给 RP2040 触发对应舵机动画
|
||
uart_send_string(emotion);
|
||
display->SetChatMessage("system", message);
|
||
if (!sound.empty()) {
|
||
audio_service_.PlaySound(sound);
|
||
}
|
||
}
|
||
|
||
void Application::DismissAlert() {
|
||
if (device_state_ == kDeviceStateIdle) {
|
||
auto display = Board::GetInstance().GetDisplay();
|
||
display->SetStatus(Lang::Strings::STANDBY);
|
||
display->SetEmotion("neutral");
|
||
uart_send_string("idle");
|
||
display->SetChatMessage("system", "");
|
||
}
|
||
}
|
||
|
||
void Application::ToggleChatState() {
|
||
if (device_state_ == kDeviceStateActivating) {
|
||
SetDeviceState(kDeviceStateIdle);
|
||
return;
|
||
} else if (device_state_ == kDeviceStateWifiConfiguring) {
|
||
audio_service_.EnableAudioTesting(true);
|
||
SetDeviceState(kDeviceStateAudioTesting);
|
||
return;
|
||
} else if (device_state_ == kDeviceStateAudioTesting) {
|
||
audio_service_.EnableAudioTesting(false);
|
||
SetDeviceState(kDeviceStateWifiConfiguring);
|
||
return;
|
||
}
|
||
|
||
if (!protocol_) {
|
||
ESP_LOGE(TAG, "Protocol not initialized");
|
||
return;
|
||
}
|
||
|
||
if (device_state_ == kDeviceStateIdle) {
|
||
Schedule([this]() {
|
||
if (!protocol_->IsAudioChannelOpened()) {
|
||
SetDeviceState(kDeviceStateConnecting);
|
||
if (!protocol_->OpenAudioChannel()) {
|
||
return;
|
||
}
|
||
}
|
||
|
||
SetListeningMode(aec_mode_ == kAecOff ? kListeningModeAutoStop : kListeningModeRealtime);
|
||
});
|
||
} else if (device_state_ == kDeviceStateSpeaking) {
|
||
Schedule([this]() {
|
||
AbortSpeaking(kAbortReasonNone);
|
||
});
|
||
} else if (device_state_ == kDeviceStateListening) {
|
||
Schedule([this]() {
|
||
protocol_->CloseAudioChannel();
|
||
});
|
||
}
|
||
}
|
||
|
||
void Application::StartListening() {
|
||
if (device_state_ == kDeviceStateActivating) {
|
||
SetDeviceState(kDeviceStateIdle);
|
||
return;
|
||
} else if (device_state_ == kDeviceStateWifiConfiguring) {
|
||
audio_service_.EnableAudioTesting(true);
|
||
SetDeviceState(kDeviceStateAudioTesting);
|
||
return;
|
||
}
|
||
|
||
if (!protocol_) {
|
||
ESP_LOGE(TAG, "Protocol not initialized");
|
||
return;
|
||
}
|
||
|
||
if (device_state_ == kDeviceStateIdle) {
|
||
Schedule([this]() {
|
||
if (!protocol_->IsAudioChannelOpened()) {
|
||
SetDeviceState(kDeviceStateConnecting);
|
||
if (!protocol_->OpenAudioChannel()) {
|
||
return;
|
||
}
|
||
}
|
||
|
||
SetListeningMode(kListeningModeManualStop);
|
||
});
|
||
} else if (device_state_ == kDeviceStateSpeaking) {
|
||
Schedule([this]() {
|
||
AbortSpeaking(kAbortReasonNone);
|
||
SetListeningMode(kListeningModeManualStop);
|
||
});
|
||
}
|
||
}
|
||
|
||
void Application::StopListening() {
|
||
if (device_state_ == kDeviceStateAudioTesting) {
|
||
audio_service_.EnableAudioTesting(false);
|
||
SetDeviceState(kDeviceStateWifiConfiguring);
|
||
return;
|
||
}
|
||
|
||
const std::array<int, 3> valid_states = {
|
||
kDeviceStateListening,
|
||
kDeviceStateSpeaking,
|
||
kDeviceStateIdle,
|
||
};
|
||
// If not valid, do nothing
|
||
if (std::find(valid_states.begin(), valid_states.end(), device_state_) == valid_states.end()) {
|
||
return;
|
||
}
|
||
|
||
Schedule([this]() {
|
||
if (device_state_ == kDeviceStateListening) {
|
||
protocol_->SendStopListening();
|
||
SetDeviceState(kDeviceStateIdle);
|
||
}
|
||
});
|
||
}
|
||
|
||
void Application::Start() {
|
||
auto& board = Board::GetInstance();
|
||
SetDeviceState(kDeviceStateStarting);
|
||
|
||
/* Setup the display */
|
||
auto display = board.GetDisplay();
|
||
|
||
// Print board name/version info
|
||
display->SetChatMessage("system", SystemInfo::GetUserAgent().c_str());
|
||
|
||
/* Setup the audio service */
|
||
auto codec = board.GetAudioCodec();
|
||
audio_service_.Initialize(codec);
|
||
audio_service_.Start();
|
||
|
||
AudioServiceCallbacks callbacks;
|
||
callbacks.on_send_queue_available = [this]() {
|
||
xEventGroupSetBits(event_group_, MAIN_EVENT_SEND_AUDIO);
|
||
};
|
||
callbacks.on_wake_word_detected = [this](const std::string& wake_word) {
|
||
xEventGroupSetBits(event_group_, MAIN_EVENT_WAKE_WORD_DETECTED);
|
||
};
|
||
callbacks.on_vad_change = [this](bool speaking) {
|
||
xEventGroupSetBits(event_group_, MAIN_EVENT_VAD_CHANGE);
|
||
};
|
||
audio_service_.SetCallbacks(callbacks);
|
||
|
||
// Start the main event loop task with priority 3
|
||
xTaskCreate([](void* arg) {
|
||
((Application*)arg)->MainEventLoop();
|
||
vTaskDelete(NULL);
|
||
}, "main_event_loop", 2048 * 4, this, 3, &main_event_loop_task_handle_);
|
||
|
||
/* Start the clock timer to update the status bar */
|
||
esp_timer_start_periodic(clock_timer_handle_, 1000000);
|
||
|
||
/* Wait for the network to be ready */
|
||
board.StartNetwork();
|
||
|
||
// Update the status bar immediately to show the network state
|
||
display->UpdateStatusBar(true);
|
||
|
||
// Check for new assets version
|
||
CheckAssetsVersion();
|
||
|
||
// Check for new firmware version or get the MQTT broker address
|
||
Ota ota;
|
||
CheckNewVersion(ota);
|
||
|
||
// Initialize the protocol
|
||
display->SetStatus(Lang::Strings::LOADING_PROTOCOL);
|
||
|
||
// Add MCP common tools before initializing the protocol
|
||
auto& mcp_server = McpServer::GetInstance();
|
||
mcp_server.AddCommonTools();
|
||
mcp_server.AddUserOnlyTools();
|
||
|
||
if (ota.HasMqttConfig()) {
|
||
protocol_ = std::make_unique<MqttProtocol>();
|
||
} else if (ota.HasWebsocketConfig()) {
|
||
protocol_ = std::make_unique<WebsocketProtocol>();
|
||
} else {
|
||
ESP_LOGW(TAG, "No protocol specified in the OTA config, using MQTT");
|
||
protocol_ = std::make_unique<MqttProtocol>();
|
||
}
|
||
|
||
protocol_->OnConnected([this]() {
|
||
DismissAlert();
|
||
});
|
||
|
||
protocol_->OnNetworkError([this](const std::string& message) {
|
||
last_error_message_ = message;
|
||
xEventGroupSetBits(event_group_, MAIN_EVENT_ERROR);
|
||
});
|
||
protocol_->OnIncomingAudio([this](std::unique_ptr<AudioStreamPacket> packet) {
|
||
if (device_state_ == kDeviceStateSpeaking) {
|
||
audio_service_.PushPacketToDecodeQueue(std::move(packet));
|
||
}
|
||
});
|
||
protocol_->OnAudioChannelOpened([this, codec, &board]() {
|
||
board.SetPowerSaveMode(false);
|
||
if (protocol_->server_sample_rate() != codec->output_sample_rate()) {
|
||
ESP_LOGW(TAG, "Server sample rate %d does not match device output sample rate %d, resampling may cause distortion",
|
||
protocol_->server_sample_rate(), codec->output_sample_rate());
|
||
}
|
||
});
|
||
protocol_->OnAudioChannelClosed([this, &board]() {
|
||
board.SetPowerSaveMode(true);
|
||
Schedule([this]() {
|
||
auto display = Board::GetInstance().GetDisplay();
|
||
display->SetChatMessage("system", "");
|
||
SetDeviceState(kDeviceStateIdle);
|
||
});
|
||
});
|
||
protocol_->OnIncomingJson([this, display](const cJSON* root) {
|
||
// Parse JSON data
|
||
auto type = cJSON_GetObjectItem(root, "type");
|
||
if (strcmp(type->valuestring, "tts") == 0) {
|
||
auto state = cJSON_GetObjectItem(root, "state");
|
||
if (strcmp(state->valuestring, "start") == 0) {
|
||
Schedule([this]() {
|
||
aborted_ = false;
|
||
if (device_state_ == kDeviceStateIdle || device_state_ == kDeviceStateListening) {
|
||
SetDeviceState(kDeviceStateSpeaking);
|
||
}
|
||
});
|
||
} else if (strcmp(state->valuestring, "stop") == 0) {
|
||
Schedule([this]() {
|
||
if (device_state_ == kDeviceStateSpeaking) {
|
||
if (listening_mode_ == kListeningModeManualStop) {
|
||
SetDeviceState(kDeviceStateIdle);
|
||
} else {
|
||
SetDeviceState(kDeviceStateListening);
|
||
}
|
||
}
|
||
});
|
||
} else if (strcmp(state->valuestring, "sentence_start") == 0) {
|
||
auto text = cJSON_GetObjectItem(root, "text");
|
||
if (cJSON_IsString(text)) {
|
||
ESP_LOGI(TAG, "<< %s", text->valuestring);
|
||
Schedule([this, display, message = std::string(text->valuestring)]() {
|
||
display->SetChatMessage("assistant", message.c_str());
|
||
});
|
||
}
|
||
}
|
||
} else if (strcmp(type->valuestring, "stt") == 0) {
|
||
auto text = cJSON_GetObjectItem(root, "text");
|
||
if (cJSON_IsString(text)) {
|
||
ESP_LOGI(TAG, ">> %s", text->valuestring);
|
||
Schedule([this, display, message = std::string(text->valuestring)]() {
|
||
display->SetChatMessage("user", message.c_str());
|
||
});
|
||
}
|
||
} else if (strcmp(type->valuestring, "llm") == 0) {
|
||
// LLM 回复中携带的情绪标签（如 "happy"、"thinking"）
|
||
auto emotion = cJSON_GetObjectItem(root, "emotion");
|
||
if (cJSON_IsString(emotion)) {
|
||
Schedule([this, display, emotion_str = std::string(emotion->valuestring)]() {
|
||
display->SetEmotion(emotion_str.c_str());
|
||
// 将 AI 返回的情绪标签发送给 RP2040，实时切换舵机动画表情
|
||
uart_send_string(emotion_str.c_str());
|
||
});
|
||
}
|
||
} else if (strcmp(type->valuestring, "mcp") == 0) {
|
||
auto payload = cJSON_GetObjectItem(root, "payload");
|
||
if (cJSON_IsObject(payload)) {
|
||
McpServer::GetInstance().ParseMessage(payload);
|
||
}
|
||
} else if (strcmp(type->valuestring, "system") == 0) {
|
||
auto command = cJSON_GetObjectItem(root, "command");
|
||
if (cJSON_IsString(command)) {
|
||
ESP_LOGI(TAG, "System command: %s", command->valuestring);
|
||
if (strcmp(command->valuestring, "reboot") == 0) {
|
||
// Do a reboot if user requests a OTA update
|
||
Schedule([this]() {
|
||
Reboot();
|
||
});
|
||
} else {
|
||
ESP_LOGW(TAG, "Unknown system command: %s", command->valuestring);
|
||
}
|
||
}
|
||
} else if (strcmp(type->valuestring, "alert") == 0) {
|
||
auto status = cJSON_GetObjectItem(root, "status");
|
||
auto message = cJSON_GetObjectItem(root, "message");
|
||
auto emotion = cJSON_GetObjectItem(root, "emotion");
|
||
if (cJSON_IsString(status) && cJSON_IsString(message) && cJSON_IsString(emotion)) {
|
||
Alert(status->valuestring, message->valuestring, emotion->valuestring, Lang::Sounds::OGG_VIBRATION);
|
||
} else {
|
||
ESP_LOGW(TAG, "Alert command requires status, message and emotion");
|
||
}
|
||
#if CONFIG_RECEIVE_CUSTOM_MESSAGE
|
||
} else if (strcmp(type->valuestring, "custom") == 0) {
|
||
auto payload = cJSON_GetObjectItem(root, "payload");
|
||
ESP_LOGI(TAG, "Received custom message: %s", cJSON_PrintUnformatted(root));
|
||
if (cJSON_IsObject(payload)) {
|
||
Schedule([this, display, payload_str = std::string(cJSON_PrintUnformatted(payload))]() {
|
||
display->SetChatMessage("system", payload_str.c_str());
|
||
});
|
||
} else {
|
||
ESP_LOGW(TAG, "Invalid custom message format: missing payload");
|
||
}
|
||
#endif
|
||
} else {
|
||
ESP_LOGW(TAG, "Unknown message type: %s", type->valuestring);
|
||
}
|
||
});
|
||
bool protocol_started = protocol_->Start();
|
||
|
||
// [T01] 摄像头 V4L2 原始采集 sanity probe（Phase 01 验证 OV3660 底层链路）
|
||
// 完成 T04 CaptureForDetection 后删除此段调用（保留 ProbeFrameCapture API 作诊断用）
|
||
#ifndef CONFIG_IDF_TARGET_ESP32
|
||
{
|
||
auto* cam = dynamic_cast<Esp32Camera*>(board.GetCamera());
|
||
if (cam) {
|
||
int64_t elapsed = 0;
|
||
bool ok = cam->ProbeFrameCapture(&elapsed);
|
||
ESP_LOGI("T01_Probe", "V4L2 probe result=%d elapsed=%lldus",
|
||
ok, (long long)elapsed);
|
||
} else {
|
||
ESP_LOGW("T01_Probe", "no camera instance (board.GetCamera() returned null or non-Esp32Camera)");
|
||
}
|
||
}
|
||
#endif
|
||
|
||
SystemInfo::PrintHeapStats();
|
||
SetDeviceState(kDeviceStateIdle);
|
||
|
||
has_server_time_ = ota.HasServerTime();
|
||
if (protocol_started) {
|
||
std::string message = std::string(Lang::Strings::VERSION) + ota.GetCurrentVersion();
|
||
display->ShowNotification(message.c_str());
|
||
display->SetChatMessage("system", "");
|
||
// Play the success sound to indicate the device is ready
|
||
audio_service_.PlaySound(Lang::Sounds::OGG_SUCCESS);
|
||
}
|
||
}
|
||
|
||
// Add a async task to MainLoop
|
||
void Application::Schedule(std::function<void()> callback) {
|
||
{
|
||
std::lock_guard<std::mutex> lock(mutex_);
|
||
main_tasks_.push_back(std::move(callback));
|
||
}
|
||
xEventGroupSetBits(event_group_, MAIN_EVENT_SCHEDULE);
|
||
}
|
||
|
||
// The Main Event Loop controls the chat state and websocket connection
|
||
// If other tasks need to access the websocket or chat state,
|
||
// they should use Schedule to call this function
|
||
void Application::MainEventLoop() {
|
||
while (true) {
|
||
auto bits = xEventGroupWaitBits(event_group_, MAIN_EVENT_SCHEDULE |
|
||
MAIN_EVENT_SEND_AUDIO |
|
||
MAIN_EVENT_WAKE_WORD_DETECTED |
|
||
MAIN_EVENT_VAD_CHANGE |
|
||
MAIN_EVENT_CLOCK_TICK |
|
||
MAIN_EVENT_ERROR, pdTRUE, pdFALSE, portMAX_DELAY);
|
||
|
||
if (bits & MAIN_EVENT_ERROR) {
|
||
SetDeviceState(kDeviceStateIdle);
|
||
Alert(Lang::Strings::ERROR, last_error_message_.c_str(), "circle_xmark", Lang::Sounds::OGG_EXCLAMATION);
|
||
}
|
||
|
||
if (bits & MAIN_EVENT_SEND_AUDIO) {
|
||
while (auto packet = audio_service_.PopPacketFromSendQueue()) {
|
||
if (protocol_ && !protocol_->SendAudio(std::move(packet))) {
|
||
break;
|
||
}
|
||
}
|
||
}
|
||
|
||
if (bits & MAIN_EVENT_WAKE_WORD_DETECTED) {
|
||
OnWakeWordDetected();
|
||
}
|
||
|
||
if (bits & MAIN_EVENT_VAD_CHANGE) {
|
||
if (device_state_ == kDeviceStateListening) {
|
||
auto led = Board::GetInstance().GetLed();
|
||
led->OnStateChanged();
|
||
}
|
||
}
|
||
|
||
if (bits & MAIN_EVENT_SCHEDULE) {
|
||
std::unique_lock<std::mutex> lock(mutex_);
|
||
auto tasks = std::move(main_tasks_);
|
||
lock.unlock();
|
||
for (auto& task : tasks) {
|
||
task();
|
||
}
|
||
}
|
||
|
||
if (bits & MAIN_EVENT_CLOCK_TICK) {
|
||
clock_ticks_++;
|
||
auto display = Board::GetInstance().GetDisplay();
|
||
display->UpdateStatusBar();
|
||
|
||
// Print the debug info every 10 seconds
|
||
if (clock_ticks_ % 10 == 0) {
|
||
// SystemInfo::PrintTaskCpuUsage(pdMS_TO_TICKS(1000));
|
||
// SystemInfo::PrintTaskList();
|
||
SystemInfo::PrintHeapStats();
|
||
}
|
||
}
|
||
}
|
||
}
|
||
|
||
void Application::OnWakeWordDetected() {
|
||
if (!protocol_) {
|
||
return;
|
||
}
|
||
|
||
if (device_state_ == kDeviceStateIdle) {
|
||
audio_service_.EncodeWakeWord();
|
||
|
||
if (!protocol_->IsAudioChannelOpened()) {
|
||
SetDeviceState(kDeviceStateConnecting);
|
||
if (!protocol_->OpenAudioChannel()) {
|
||
audio_service_.EnableWakeWordDetection(true);
|
||
return;
|
||
}
|
||
}
|
||
|
||
auto wake_word = audio_service_.GetLastWakeWord();
|
||
ESP_LOGI(TAG, "Wake word detected: %s", wake_word.c_str());
|
||
#if CONFIG_SEND_WAKE_WORD_DATA
|
||
// Encode and send the wake word data to the server
|
||
while (auto packet = audio_service_.PopWakeWordPacket()) {
|
||
protocol_->SendAudio(std::move(packet));
|
||
}
|
||
// Set the chat state to wake word detected
|
||
protocol_->SendWakeWordDetected(wake_word);
|
||
SetListeningMode(aec_mode_ == kAecOff ? kListeningModeAutoStop : kListeningModeRealtime);
|
||
#else
|
||
SetListeningMode(aec_mode_ == kAecOff ? kListeningModeAutoStop : kListeningModeRealtime);
|
||
// Play the pop up sound to indicate the wake word is detected
|
||
audio_service_.PlaySound(Lang::Sounds::OGG_POPUP);
|
||
#endif
|
||
} else if (device_state_ == kDeviceStateSpeaking) {
|
||
AbortSpeaking(kAbortReasonWakeWordDetected);
|
||
} else if (device_state_ == kDeviceStateActivating) {
|
||
SetDeviceState(kDeviceStateIdle);
|
||
}
|
||
}
|
||
|
||
void Application::AbortSpeaking(AbortReason reason) {
|
||
ESP_LOGI(TAG, "Abort speaking");
|
||
aborted_ = true;
|
||
if (protocol_) {
|
||
protocol_->SendAbortSpeaking(reason);
|
||
}
|
||
}
|
||
|
||
void Application::SetListeningMode(ListeningMode mode) {
|
||
listening_mode_ = mode;
|
||
SetDeviceState(kDeviceStateListening);
|
||
}
|
||
|
||
void Application::SetDeviceState(DeviceState state) {
|
||
if (device_state_ == state) {
|
||
return;
|
||
}
|
||
|
||
clock_ticks_ = 0;
|
||
auto previous_state = device_state_;
|
||
device_state_ = state;
|
||
ESP_LOGI(TAG, "STATE: %s", STATE_STRINGS[device_state_]);
|
||
|
||
// Send the state change event
|
||
DeviceStateEventManager::GetInstance().PostStateChangeEvent(previous_state, state);
|
||
|
||
auto& board = Board::GetInstance();
|
||
auto display = board.GetDisplay();
|
||
auto led = board.GetLed();
|
||
led->OnStateChanged();
|
||
// 根据设备状态通过 UART 发送指令给 RP2040，驱动舵机执行对应动画
|
||
switch (state) {
|
||
case kDeviceStateUnknown:
|
||
case kDeviceStateIdle:
|
||
display->SetStatus(Lang::Strings::STANDBY);
|
||
display->SetEmotion("neutral");
|
||
audio_service_.EnableVoiceProcessing(false);
|
||
audio_service_.EnableWakeWordDetection(true);
|
||
uart_send_string("idle"); // → RP2040: 进入睡眠姿态（闭眼、耳朵收起）
|
||
break;
|
||
case kDeviceStateConnecting:
|
||
display->SetStatus(Lang::Strings::CONNECTING);
|
||
display->SetEmotion("neutral");
|
||
display->SetChatMessage("system", "");
|
||
break;
|
||
case kDeviceStateListening:
|
||
display->SetStatus(Lang::Strings::LISTENING);
|
||
display->SetEmotion("neutral");
|
||
uart_send_string("listening"); // → RP2040: 进入聆听姿态（好奇歪头、眼球随机看）
|
||
|
||
// Make sure the audio processor is running
|
||
if (!audio_service_.IsAudioProcessorRunning()) {
|
||
// Send the start listening command
|
||
protocol_->SendStartListening(listening_mode_);
|
||
audio_service_.EnableVoiceProcessing(true);
|
||
audio_service_.EnableWakeWordDetection(false);
|
||
}
|
||
break;
|
||
case kDeviceStateSpeaking:
|
||
display->SetStatus(Lang::Strings::SPEAKING);
|
||
uart_send_string("speaking"); // → RP2040: 进入说话姿态（嘴巴开合、眼球随机看、身体摆动）
|
||
if (listening_mode_ != kListeningModeRealtime) {
|
||
audio_service_.EnableVoiceProcessing(false);
|
||
// Only AFE wake word can be detected in speaking mode
|
||
audio_service_.EnableWakeWordDetection(audio_service_.IsAfeWakeWord());
|
||
}
|
||
audio_service_.ResetDecoder();
|
||
break;
|
||
default:
|
||
// Do nothing
|
||
break;
|
||
}
|
||
}
|
||
|
||
void Application::Reboot() {
|
||
ESP_LOGI(TAG, "Rebooting...");
|
||
// Disconnect the audio channel
|
||
if (protocol_ && protocol_->IsAudioChannelOpened()) {
|
||
protocol_->CloseAudioChannel();
|
||
}
|
||
protocol_.reset();
|
||
audio_service_.Stop();
|
||
|
||
vTaskDelay(pdMS_TO_TICKS(1000));
|
||
esp_restart();
|
||
}
|
||
|
||
bool Application::UpgradeFirmware(Ota& ota, const std::string& url) {
|
||
auto& board = Board::GetInstance();
|
||
auto display = board.GetDisplay();
|
||
|
||
// Use provided URL or get from OTA object
|
||
std::string upgrade_url = url.empty() ? ota.GetFirmwareUrl() : url;
|
||
std::string version_info = url.empty() ? ota.GetFirmwareVersion() : "(Manual upgrade)";
|
||
|
||
// Close audio channel if it's open
|
||
if (protocol_ && protocol_->IsAudioChannelOpened()) {
|
||
ESP_LOGI(TAG, "Closing audio channel before firmware upgrade");
|
||
protocol_->CloseAudioChannel();
|
||
}
|
||
ESP_LOGI(TAG, "Starting firmware upgrade from URL: %s", upgrade_url.c_str());
|
||
|
||
Alert(Lang::Strings::OTA_UPGRADE, Lang::Strings::UPGRADING, "download", Lang::Sounds::OGG_UPGRADE);
|
||
vTaskDelay(pdMS_TO_TICKS(3000));
|
||
|
||
SetDeviceState(kDeviceStateUpgrading);
|
||
|
||
std::string message = std::string(Lang::Strings::NEW_VERSION) + version_info;
|
||
display->SetChatMessage("system", message.c_str());
|
||
|
||
board.SetPowerSaveMode(false);
|
||
audio_service_.Stop();
|
||
vTaskDelay(pdMS_TO_TICKS(1000));
|
||
|
||
bool upgrade_success = ota.StartUpgradeFromUrl(upgrade_url, [display](int progress, size_t speed) {
|
||
std::thread([display, progress, speed]() {
|
||
char buffer[32];
|
||
snprintf(buffer, sizeof(buffer), "%d%% %uKB/s", progress, speed / 1024);
|
||
display->SetChatMessage("system", buffer);
|
||
}).detach();
|
||
});
|
||
|
||
if (!upgrade_success) {
|
||
// Upgrade failed, restart audio service and continue running
|
||
ESP_LOGE(TAG, "Firmware upgrade failed, restarting audio service and continuing operation...");
|
||
audio_service_.Start(); // Restart audio service
|
||
board.SetPowerSaveMode(true); // Restore power save mode
|
||
Alert(Lang::Strings::ERROR, Lang::Strings::UPGRADE_FAILED, "circle_xmark", Lang::Sounds::OGG_EXCLAMATION);
|
||
vTaskDelay(pdMS_TO_TICKS(3000));
|
||
return false;
|
||
} else {
|
||
// Upgrade success, reboot immediately
|
||
ESP_LOGI(TAG, "Firmware upgrade successful, rebooting...");
|
||
display->SetChatMessage("system", "Upgrade successful, rebooting...");
|
||
vTaskDelay(pdMS_TO_TICKS(1000)); // Brief pause to show message
|
||
Reboot();
|
||
return true;
|
||
}
|
||
}
|
||
|
||
void Application::WakeWordInvoke(const std::string& wake_word) {
|
||
if (!protocol_) {
|
||
return;
|
||
}
|
||
|
||
if (device_state_ == kDeviceStateIdle) {
|
||
audio_service_.EncodeWakeWord();
|
||
|
||
if (!protocol_->IsAudioChannelOpened()) {
|
||
SetDeviceState(kDeviceStateConnecting);
|
||
if (!protocol_->OpenAudioChannel()) {
|
||
audio_service_.EnableWakeWordDetection(true);
|
||
return;
|
||
}
|
||
}
|
||
|
||
ESP_LOGI(TAG, "Wake word detected: %s", wake_word.c_str());
|
||
#if CONFIG_USE_AFE_WAKE_WORD || CONFIG_USE_CUSTOM_WAKE_WORD
|
||
// Encode and send the wake word data to the server
|
||
while (auto packet = audio_service_.PopWakeWordPacket()) {
|
||
protocol_->SendAudio(std::move(packet));
|
||
}
|
||
// Set the chat state to wake word detected
|
||
protocol_->SendWakeWordDetected(wake_word);
|
||
SetListeningMode(aec_mode_ == kAecOff ? kListeningModeAutoStop : kListeningModeRealtime);
|
||
#else
|
||
SetListeningMode(aec_mode_ == kAecOff ? kListeningModeAutoStop : kListeningModeRealtime);
|
||
// Play the pop up sound to indicate the wake word is detected
|
||
audio_service_.PlaySound(Lang::Sounds::OGG_POPUP);
|
||
#endif
|
||
} else if (device_state_ == kDeviceStateSpeaking) {
|
||
Schedule([this]() {
|
||
AbortSpeaking(kAbortReasonNone);
|
||
});
|
||
} else if (device_state_ == kDeviceStateListening) {
|
||
Schedule([this]() {
|
||
if (protocol_) {
|
||
protocol_->CloseAudioChannel();
|
||
}
|
||
});
|
||
}
|
||
}
|
||
|
||
bool Application::CanEnterSleepMode() {
|
||
if (device_state_ != kDeviceStateIdle) {
|
||
return false;
|
||
}
|
||
|
||
if (protocol_ && protocol_->IsAudioChannelOpened()) {
|
||
return false;
|
||
}
|
||
|
||
if (!audio_service_.IsIdle()) {
|
||
return false;
|
||
}
|
||
|
||
// Now it is safe to enter sleep mode
|
||
return true;
|
||
}
|
||
|
||
void Application::SendMcpMessage(const std::string& payload) {
|
||
if (protocol_ == nullptr) {
|
||
return;
|
||
}
|
||
|
||
// Make sure you are using main thread to send MCP message
|
||
if (xTaskGetCurrentTaskHandle() == main_event_loop_task_handle_) {
|
||
protocol_->SendMcpMessage(payload);
|
||
} else {
|
||
Schedule([this, payload = std::move(payload)]() {
|
||
protocol_->SendMcpMessage(payload);
|
||
});
|
||
}
|
||
}
|
||
|
||
void Application::SetAecMode(AecMode mode) {
|
||
aec_mode_ = mode;
|
||
Schedule([this]() {
|
||
auto& board = Board::GetInstance();
|
||
auto display = board.GetDisplay();
|
||
switch (aec_mode_) {
|
||
case kAecOff:
|
||
audio_service_.EnableDeviceAec(false);
|
||
display->ShowNotification(Lang::Strings::RTC_MODE_OFF);
|
||
break;
|
||
case kAecOnServerSide:
|
||
audio_service_.EnableDeviceAec(false);
|
||
display->ShowNotification(Lang::Strings::RTC_MODE_ON);
|
||
break;
|
||
case kAecOnDeviceSide:
|
||
audio_service_.EnableDeviceAec(true);
|
||
display->ShowNotification(Lang::Strings::RTC_MODE_ON);
|
||
break;
|
||
}
|
||
|
||
// If the AEC mode is changed, close the audio channel
|
||
if (protocol_ && protocol_->IsAudioChannelOpened()) {
|
||
protocol_->CloseAudioChannel();
|
||
}
|
||
});
|
||
}
|
||
|
||
void Application::PlaySound(const std::string_view& sound) {
|
||
audio_service_.PlaySound(sound);
|
||
} |