Skip to content

Latest commit

 

History

History
151 lines (137 loc) · 7.15 KB

README_TODO_002.md

File metadata and controls

151 lines (137 loc) · 7.15 KB

ESP32, voice recording

tensorflow lite micro (tflite) esp32 port

tensorflow lite micro (tflite) stm32 / mbed / arm port / etc

TFLite work

  • blink_v2_micro_speech_success.tar.gz
    with esp-idf-v3.3.4
  • freertos_stm32f103rct6_v3_queue.rar
    STM32 CMSIS-FreeRTOS demo
  • freertos_stm32f103rct6_v4_ac6.rar
    STM32 CMSIS-FreeRTOS demo for ac6
  • inmp441_stm32f411re_v2_success_3bit4bit.rar
    stm32f411re inmp441 i2s
  • WM8960_Record_v3_success_inmp441.rar
    stm32f103ze inmp441 i2s
  • WM8960_Record_stm32f103ze_compare.rar
    stm32f103ze inmp441 i2s mod from WM8960_Audio_Board_Code, differences
  • microspeech_stm32f411re_v4_compile.rar
    final stm32 project micro_speech ac6
  • micro_speech_vs2013_success.rar
    win32, vs2013
  • microspeech_stm32h743vi_v4.rar
    stm32h743 project micro_speech ac6 not tested

TFLite work TODO

ML-KWS, nucleo-f411re

//UART1_TX==PA_9==D8<->FT232.RXD  
//UART1_RX==PA_10==D2<->FT232.TXD  
serial_init(&stdio_uart, PA_9, PA_10);  //redirect to Serial1  
stdio_uart_inited = 1;   
printf("ready\r\n");
  • (TODO) search baidupan, init template project, blink_v1_stm32f411ce_init.rar

speech_commands, tensorflow 1.5.0

TODO: 待解决，用i5的电脑开虚拟机已经可以正常安装tf 2.x，但旧电脑好像不行

  • 命令行参照这篇：
    https://www.cnblogs.com/lijianming180/p/12258774.html
    我不是自己编译tensorflow的，我用的方法是用xubuntu安装python2（就是2.7）
    和python2-pip（需要特殊方法安装pip），然后再离线安装tensorflow 1.5
    （不过依赖包还是在线安装），简单说，就是我利用旧版本来绕过CPU指令集的问题
    （直接安装最新cpu版本，运行是会报错的）。除了安装问题，还有一个问题是训练
    模型的时间非常长，我现在没有彻底跑完整个train.py（我估计要跑一天），我在
    想有没有办法缩短训练数据的时间，或者是否存在断点执行的方法
  • search baidupan, tensorflow-1.5.0-cp27-none-linux_x86_64.whl
  • search baidupan, tensorflow-1.5.0.zip
  • search baidupan, speech_commands_v0.01.tar.gz
  • 安装Python 2和pip2:
    see https://www.cnblogs.com/zhuangliu/archive/2016/11/20/6083063.html
    (???) $ sudo apt-get install python2.7
    $ wget https://bootstrap.pypa.io/get-pip.py
    $ sudo python2 get-pip.py
    $ sudo python2 -m pip install tensorflow-1.5.0-cp27-none-linux_x86_64.whl
    $ python2
  • other study project
    (TODO) https://github.com/accraze/keyword-spotter
    search here, keyword-spotter

(TODO, baidupan) stm32 sound record

  • (TODO, in home computer) stm32f103zet6, Open103Z_I2S, WM8960_Record, WM8960_Audio_Board_Code_v1.rar
  • STM32F4-Discovery_FW_V1.1.0, Audio_playback_and_record, en.stsw-stm32068.zip
  • stm32F4_dsp_microphone_fft_rtos.rar
  • (TODO) STM32F4 i2s/adc/pdm recording code
  • ref https://os.mbed.com/code/

(TODO, baidupan) stm32 (and other) sound process

  • STM32F407VG, ASR_Project.rar
  • STM32F407VG, Design_Project-Speech_Recognition_on_Embedded_System.rar
  • k210, m5stickv-tensorflow-lite-micro.rar, maixcube-tensorflow-lite-micro.rar
  • esp32, ML-KWS-for-ESP32.rar
  • STM32F746NG, ml-kws-for-mcu_alxkbr.rar
  • STM32F407VGT6_1, SmartPillow.rar
  • STM32F429ZI, Speaker-Recognition-System-in-ARM.rar, SR_stm32f429zi.rar
  • STM32F407VG, Voice-Recognition.rar
  • mic_vad_streaming, DeepSpeech-examples.rar
  • (TODO) STM32f103VE, stm32-speech-recognition_v2.rar

TensorFlow example speech_commands

Wav analyzing Software

ADC

https://github.com/sparkfun/MEMS_Mic_Breakout-ADMP401/blob/V_1.3/Firmware/SparkFun_ADMP401_Simple_Sketch/SparkFun_INMP401.ino
https://os.mbed.com/users/rayxke/notebook/sparkfun-mems-microphone-breakout---inmp401-admp40/
https://github.com/jenfoxbot/MEMSMicHookUpGuide/blob/master/ExampleCode.ino
https://learn.sparkfun.com/tutorials/mems-microphone-hookup-guide
Arduino+MAX9814制作简易录音机
https://www.cnblogs.com/Ray-liang/p/9786154.html