A set of commands will be placed inside of a random article. The article will be read and played on a text to speech website to test if there are any false recognition.
ZHANG, Jun
January 19, 2022 3:21 pm
Can you elaborate on the “internet connection” part? What is the major difficulty?
Transfer learning is to keep improving AI recognition by using daily voice data. The voice command will be saved in the local storage. The system will send the local voice command data through the internet from time to time.
How do you evaluate the accuracy of your designed instrument. Is there any measurement results?
A set of commands will be placed inside of a random article. The article will be read and played on a text to speech website to test if there are any false recognition.
Can you elaborate on the “internet connection” part? What is the major difficulty?
Transfer learning is to keep improving AI recognition by using daily voice data. The voice command will be saved in the local storage. The system will send the local voice command data through the internet from time to time.
Similar to IoT, wireless network coverage will be the major concern in that environment.
What speech regonition and control methods will you use?
It is an offline google teachable machine AI voice recognition with I2C/UART control
Does it mean that this project will not involve anything about machine learning/AI algorithm design and training?
The AI algorithm will be adjusted if the offline recognition is not accurate enough