- Update esp-sr lib to release/v1.0.
- Fix model copy issue on Windows.
- Add the API of getting phrase ID if command word is detected.
- Currently, ESP32-S3-BOX does not support capturing DAC output through
ADC. The next version of ESP32-S3-BOX Lite will use software capture.
Therefore, ES7210 will only use ADC channels 1 and 2, and output
these 2 channels to SDOUT1. In this case, when esp-idf still has
the problem of I2S TDM mode, ES8311 will work well on I2S Philips
standard. This will also reduce memory consumption.
- The previous wake-up audio was in 32-bit dual-channel WAV format.
Since the WAV header is aligned with 4 bytes, and the WAV header is
really short, it is possible to play the WAV file directly from the
beginning. This version changes the audio file to PCM raw data with
16 bits per channel.
This merge request contains the following:
- Update the animation of the speech recognition interface to improve response speed and fluency
- Fixed the issue that the voice wake-up animation was extremely stuck when the prompt message was not closed
- Fixed the problem that the control panel information was not synchronized after the LED status was modified by voice
- Fixed the problem that the LED does not turn on when the brightness is updated through the Web
- Updated LVGL to V8.1
- Dark lines appear in the shadow of the control in V8.0 : LED with custom style
- All third-party components are moved to the LVGL epository. Remove all lv_lib_xxx and ntrol whether the component is enabled through Kconfig.
- Split file system support to stdio, POSIX and FatFS. Select stdio according to IDF support and specify the default drive letter in CMakeLists.txt
- Fixed an issue where a command word that was too short was added successfully on the web page but esp-sr refused to add it
In addition, update the factory firmware version number to V0.2.1.
icing@tomato:~#
IAMLIUBO
marcus_xu
Wu Jian Gang
Mahavir Jain
Chris Morgan
zhouli
Yu Zhe
Li Bo