1. 进一步拓展了语音识别，目前可以从sd卡导入模型，避免了model文件占用flash分区大小

2. 进一步修改了语音识别，关闭了关键词唤醒功能，只保留了指令识别功能 3. 构建了业务层的基本框架(增加了底层驱动对于的C++兼容)，业务代码采用C++编写，启用了RTTI(运行时类型识别)
2025-09-03 00:19:14 +08:00
parent ce0998c1c6
commit 5d79f88918
42 changed files with 822 additions and 100 deletions
@@ -539,8 +539,8 @@ CONFIG_PARTITION_TABLE_MD5=y
 #
 # ESP Speech Recognition
 #
-CONFIG_MODEL_IN_FLASH=y
-# CONFIG_MODEL_IN_SDCARD is not set
+# CONFIG_MODEL_IN_FLASH is not set
+CONFIG_MODEL_IN_SDCARD=y
 CONFIG_USE_AFE=y
 CONFIG_AFE_INTERFACE_V1=y
 # CONFIG_USE_NSNET is not set