Oopenvela-robotai_conversation:add timeout

9e3d17fc创建于 4月20日历史提交

文件	最后提交记录	最后更新时间
img	feat:ai_conversation - add get_system_volume function to retrieve the value from kvdb - modify mcp_get_tool_list - modify media_player_start start time - add timeout detection sercer response	2 个月前
include	feat:ai_conversation - add get_system_volume function to retrieve the value from kvdb - modify mcp_get_tool_list - modify media_player_start start time - add timeout detection sercer response	2 个月前
res	feat:ai_conversation - add get_system_volume function to retrieve the value from kvdb - modify mcp_get_tool_list - modify media_player_start start time - add timeout detection sercer response	2 个月前
src	ai_conversation:add timeout Signed-off-by: jiayadong1 <jiayadong1@xiaomi.com>	2 个月前
utils	fix: ai_chat bug - Fixed a crash issue in the volc thread. - Added volume adjustment functionality. - Implemented null pointer checks. - Resolved logic issues in music playback. - Addressed workaround for long - running conversation problem. - Ensured proper application exit. - add volc.py for test - avoid sem deadlock - optimize the implementation of the buffer Signed-off-by: jiayadong1 <jiayadong1@xiaomi.com>	2 个月前
CMakeLists.txt	feat:ai_conversation - add get_system_volume function to retrieve the value from kvdb - modify mcp_get_tool_list - modify media_player_start start time - add timeout detection sercer response	2 个月前
Kconfig	feat:ai_conversation - add get_system_volume function to retrieve the value from kvdb - modify mcp_get_tool_list - modify media_player_start start time - add timeout detection sercer response	2 个月前
Make.defs	feat: add native app in packages/demo(ai_chat) - Create core directory structure for AI chat application - Include header file (include/), resources (res/), source code (src) and utilities (utils/) directories - developers can call ai engine through ai_conversation.h interface Signed-off-by: jiayadong1 <jiayadong1@xiaomi.com>	2 个月前
Makefile	feat:ai_conversation - add get_system_volume function to retrieve the value from kvdb - modify mcp_get_tool_list - modify media_player_start start time - add timeout detection sercer response	2 个月前
PLUGIN_DEVELOPMENT.md	feat:ai_conversation - add get_system_volume function to retrieve the value from kvdb - modify mcp_get_tool_list - modify media_player_start start time - add timeout detection sercer response	2 个月前
README.md	feat:ai_conversation - add get_system_volume function to retrieve the value from kvdb - modify mcp_get_tool_list - modify media_player_start start time - add timeout detection sercer response	2 个月前
volc.py	feat:ai_conversation - add get_system_volume function to retrieve the value from kvdb - modify mcp_get_tool_list - modify media_player_start start time - add timeout detection sercer response	2 个月前

AI Chat Demo 使用说明

Demo 演示

1. 运行环境与依赖

- 基础组件：选择 `CONFIG_AI_CONVERSATION_DEMO` 后会自动 `select` MEDIA、MEDIA_SERVER、LIBUV_EXTENSION、LIB_JSONC、CRYPTO_MBEDTLS、OPENSSL_MBEDTLS_WRAPPER、NETUTILS_LIBWEBSOCKETS 等依赖。确保存量镜像已包含相关库与驱动。

- 硬件需求：设备需具备麦克风输入（默认流名 `Capture`）和扬声器输出（默认流名 `Music`），并能访问外网。

- 资源文件：UI 依赖字体、图标、音乐资源，需放置在配置路径下（详见第 3 节）。

- Volc API Key：需向火山引擎申请实时对话 API Key，并写入配置项。

2. menuconfig 选项

2.1 主体功能

AI_CONVERSATION_DEMO=y

# AI_CONVERSATION_DEMO 有对应依赖，需要打开对应依赖
MEDIA=y
MEDIA_SERVER=y
LIBUV_EXTENSION=y
LIB_JSONC=y
CRYPTO_MBEDTLS=y
OPENSSL_MBEDTLS_WRAPPER=y
NETUTILS_LIBWEBSOCKETS=y

AI_VOLC_PLUGIN=y
VOLC_PULGIN_API_KEY=开发者自己申请的key
AI_CONVERSATION_RECORDER_STREAM=cap

2.2 媒体与 GUI 参数

配置项	说明	默认值	建议
`CONFIG_AI_CONVERSATION_RECORDER_STREAM`	麦克风流名	`"Capture"`	模拟器上使用'cap'
`CONFIG_AI_CONVERSATION_GUI_ROOT`	资源根目录	`/data`	依据挂载点调整
`CONFIG_AI_CONVERSATION_GUI_RES_PATH`	根目录下资源路径	`/res`	通常保持默认
`CONFIG_AI_CONVERSATION_GUI_FONT_PATH`	字体路径	`/fonts`	确认包含 `MiSans-Normal.ttf`
`CONFIG_AI_CONVERSATION_GUI_ICONS_PATH`	图标路径	`/icons`	需要 `app.png`
`CONFIG_AI_CONVERSATION_GUI_PRIORITY`	线程优先级	`100`	视系统负载调整
`CONFIG_AI_CONVERSATION_GUI_STACKSIZE`	栈大小	`32768`	复杂 UI 可适当增大

若保持默认路径，则字体实际路径为 /data/res/fonts/MiSans-Normal.ttf，图标为 /data/res/icons/app.png。

3. 资源准备

将字体、图标、音乐等文件同步到 CONFIG_AI_CONVERSATION_GUI_ROOT + CONFIG_AI_CONVERSATION_GUI_RES_PATH 对应的目录。
音乐文件需匹配源码中的映射：例如 res/music/daoxiang.mp3、qinghuaci.mp3、Kgezhiwang.mp3、tongzhuodeni.mp3 等。

若需新增 UI 资源，可在相同目录补充文件并修改代码里的文件名。

# 安装adb
sudo apt install android-tools-adb

# 推送资源
adb push apps/packages/demos/ai_chat/res /data/

4. 运行demo

ifup eth0
renew eth0
ai_conversation &

5. AI 对话服务插件开发

后续我们会基于插件接入更多的云端服务，想自行接入的开发者请参考该文档。

AI Chat Demo 使用说明

Demo 演示

1. 运行环境与依赖

2. menuconfig 选项

2.1 主体功能

2.2 媒体与 GUI 参数

3. 资源准备

4. 运行demo

5. AI 对话服务插件开发

火山 key 的申请

参考文档

1. 进入边缘大模型网关

2. 申请语音对话智能体的 key