Hardware components | ||||||
 |
| × | 1 | |||
| × | 1 | ||||
 |
| × | 1 | |||
 |
| × | 2 | |||
| × | 1 | ||||
 |
| × | 1 | |||
Software apps and online services | ||||||
| ||||||
| ||||||
Hand tools and fabrication machines | ||||||
 |
| |||||
 |
| |||||
Today the technology has grown really well but its has dependence too. This concept is designed to sustain our life in any climatic condition. we know that there are many voice control, smart gadgets and devices around us , all required is source of internet. It may not work when you're at certain climatic condition or out-of area. Here the idea is to use the ML based solution to reorganize the voice and take controls.
Received the board for contest , and started with basic examples from https://github.com/QuickLogic-Corp here is my getting start guide with demo on using quick feather board for motion detection with ML.
After understanding the basic concept of seniML and Analytics studio i moved on to try voice recognition witch microphone example. I tried to capture the voice samples with the help of Data Capture Lab. Also gone through many tutorials served from the SensiML and QuickLogic team. Finally decided make some voice based remote system.
Which required wireless connectivity and controls to showcase the demo. Taken voice samples for color sets "RED", "BLUE" and "GREEN" for changing mood lights over voice. In another model sampled "Open" and "Close" inputs.
I have used Analytics studio for generating Knowledge packs. Experienced problems to differentiate the noise from the actual command. But finally managed with some noise capture.
On qorc-sdk side the changes are only on UART and LED's i directed the classification out to BLE Chip (CYW20819) which i have used from Cypress.
Block Diagram
I have designed the custom board for Bluetooth with 12V CMOS lowside switches which are from Infineon. And so the final setup will looks like below.
On Bluetooth side i have used simple GATT server and Client to communicate each-other.
The other end i designed a Door with lego blocks and driven by the DC motor with Infineon H-Bridge Kit2Go board
The final setup is powered with 9V battery. The Motor is controlled with PWM out.
After all setup the final demo is completed and the design of Machining Learning based Voice remote works really well. Some times i see it does not recognize the word but still its my concept design , sill lot things to improve to bring it as product. The source code will be published here
The final working demo:
Comments