Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Chinese speech recognizer & mp3 player module demo video

Status
Not open for further replies.

mdanh2002

Member
Hi,

I purchased the LD3320 MP3 player and Chinese speech recognizer module made by icroute (www.icroute.com/) from eBay a few months ago. Its Chinese voice recognition mechanism can be initialized with the Pinyin transliterations of the Chinese text to be recognized, and the module will attempt to listen to the audio sent to its input channel to identify any voice that resembles the programmed list of Chinese words sent during initialization. MP3 playback is supported as well - with headphone/lineout (stereo) and speaker (mono) outputs.

Although the PCB that I purchased seems to have some serious issues (wrong pinout labeling and ECM microphone connected in reverse polarity), I finally got it to work. I decided to make a demo video showing the module working with a PIC microcontroller and an ST7735 128x160 16-bit color LCD to display the speech recognition results. It shows the results of the module trying to recognize proper names in Chinese (bei jing [北京], shang hai [上海], hong kong [香港], chong qing [重庆], tian an men [天安门) and other words such as "a li ba ba". A single beep means that the speech is recognized while a double beep indicates unrecognized speech. Although the speech recognition quality highly depends on the input audio, volume level and other environmental conditions, overall the detection sensitivity and selectivity seems satisfactory as can be seen from the video.

The end of the video shows the stereo playback of an MP3 song stored on the SD card - using a PAM8403 amplifier whose output is fed into two 8-ohm loudspeakers. The playback quality seems good and comparable to the VS1053 module.


Let me know your thoughts. I have yet to see a similar IC for English speech recognition.

Attached to this thread is the board (front and back).
 

Attachments

  • Board Front.jpg
    Board Front.jpg
    47.3 KB · Views: 181
  • Board_Back.jpg
    Board_Back.jpg
    121.5 KB · Views: 169
Last edited:
I wish I had that IC yesterday when a Chinese girl came to my home to sell something. She tried to speeky zee English but I could not understand most things she said. I think she tried to sell me a rented furnace, air conditioner and water heater but I explained to her that I got rid of the expensive rentals years ago when I bought my own and got government rebates.

People in Pakistan make long distance phone calls to Canada to sell "Duct cleaning". Since I get too many of these calls and they can hardly speeky zee English then I tell them that My Ducks are already nice and clean, quack quack.

My son's and daughter's cell phones have English voice recognition that works sometimes.
You say "Call John" and it dials Don.
 
Nah, the Chinese language is not all that difficult.
In China even the little children can speak it very well.

JimB
 
Well, I got this module out of curiosity from eBay - just wanted to see how a Chinese speech recognition IC works. I cannot speak Chinese either - that's why I have to test using Chinese proper names which I can pronounce. Turned out it works fairly well. However what annoyed me was the wrong pinout printed on the PCB, causing me to waste several days debugging the issue - the labels of almost all communication pins on the left connection header are reversed, RSTB should be INTB, CLK should be WRB and vice versa! (The image of the board attached in my previous post is the image after being edited with Paint to show the correct pinout)

I am not sure why there should be this issue, knowing that the VCC and GND pins are correctly labelled.

A detailed writeup, with a description of the PCB issue, and the MPLAB X source code, can be found on my blog
**broken link removed**
 
Last edited:
Thanks for the thread. I saw a couple of these too in eBay - hoping that they have the English versions. I can test this since I have known people who converse in Mandarin (mine's pretty limited). :)
 
Thanks for the thread. I saw a couple of these too in eBay - hoping that they have the English versions. I can test this since I have known people who converse in Mandarin (mine's pretty limited). :)

I could not find any reasonably-priced modules for English speech recognition. Even for English text-to-speech, companies like Sparkfun charge an arm and a leg for a TTS256 or SPO256 clone. Maybe there is no market for English speech recognition IC, that's why we could not find any such IC at cheap price.

For this module you can test with proper names like what I did without actually knowing how to speak Mandarin. The detection is quite sensitive and will pick up any speech that are close enough to the expected word. It will give you a good idea of the recognition quality. I feel great after getting this module to work - all the datasheet and documentation files are in Chinese and and I spent around 2 weeks with the help of Google Translate to get speech recognition and MP3 playback working as seen in the video.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top