This project is trying to imitate the magic mirror in the movie *SnowWhite*
The project will remain in silence when there is no one arround
When someone stands infront of it for some time, it will wake up and "talk" to him/her
When the person leaves, the mirror go back to rest
System booted
Program starts to run automatically
Monitor displaying wallpaper, webcam scaning for faces
If Face detected, wake the main method
Display footage from webcam, using squares to point out eyes, mouth and the whole face
If the face continues to stay in site(50 frames), play:"请说话"("Please speak"), microphone starts to record
Voice recognition translate record to text
Turing AI gives out the answer
Voice generation translate text from Turing to record
Play record
If the text from user contains"唱一首歌"("Please sing a song"), randomly play music
else, return to displaying wallpaper
else, return to displaying wallpaper
Raspberry Pi (I am using 3B)
Monitor
Speakers
Webcam & microphone
The installation of OpenCV is needed
Best tutorial for OpenCV installation I've ever met, huge thanks to the author
Voice generation & recognition is essential, I used it from Baidu but you can also use other sources
URL to Baidu Voice recognition
The conversation AI is used from Turing, other sources may work too.
My email adress: Aaron3963@163.com