Abstract: Music is a limitless form of art. Accessibility is an essential part of its emergence. Traditional musical instruments like piano, drum, etc., are expensive and arduous which reduces its accessibility and portability. To make it universally available and efficient, this is an attempt to demonstrate the use of an alternate medium for creating music using gestures. use of colored gloves/fingertip covers/nail paints/drumsticks, etc., making it more user-friendly and autonomous.This makes it available to a large number of music enthusiasts. Devices, nowadays, use touch-based and key-array-based controls, limiting the interface to two dimensions. For this purpose, object detection has been used to demonstrate a gesture-controlled piano, giving it another dimension. Existing image processing systems use shadow detection methods, IR detection methods, etc., whereas here, object detection method for gesture recognition has been used to identify the gesture and simulate the playing of piano notes. The object detection method eliminates the use of colored gloves/fingertip cover/nail paints/drum sticks, etc, making it more user friendly and autonomous.
Keywords: CNN, API, OpenCV, GPU, CPU
|
DOI:
10.17148/IJARCCE.2021.105160
