This post is a tad bit late so I will get right to it! I finalized my NodeBot and if you have been following along you are probably wondering.. Well? What is it?
I made (what I like to call) the Frozen NodeBot! Using an Arduino, the MP3 Music Shield and Microsoft Cognitive Services, I created a voice activated singing Elsa Doll. The inspiration behind it all, is my six year old sister who is always bugging me to play when I have to work. I had to find a way to make my work time, her play time. So why not make a doll right?
I dont actually build the doll itself, rather I bought a doll and added some hardware to it! The process is very simple:
- The “doll “listens for about 10 seconds for any voice or sound
- It records these 10 seconds and then pipes it down into a wav file
- The wav file is then sent to Microsoft Cognitive Services so that it can be translated to text using the Speech to Text api. The text is then sent back.
- The text buffer is analyzed for the keywords that will trigger the doll to sing
- If the words are found, serial port kicks in and tells the MP3 player shield to play a random track off of it’s SD card
- Voila! You have happy child singing to a 15 second snippet of the song Let It Go
See a video of it working (just the hardware) : https://youtu.be/Z1cDxRT_qw0
Keep in mind the bot does not actually have to be an Frozen based. The code can stay the same, just change the doll, keywords and music to fit the character of your child’s choosing.
If you would like to build your own or see where to start, check out my Frozen NodeBot site here ! The materials needed and the code to get started are all available on the site. If you would like to shoot directly to the code, here you go!