Similar Music Instrument Detection via Deep Convolution YOLO-Generative Adversarial Network
Christine Dewi, Rung-Ching Chen, Hendry Hendry, Yanting Liu · 2019
Object detection and image recognition are important research topics in machine learning and artificial intelligence. The major challenge of the computer vision image recognition is to detect and recognize a similar object. Generative Adversarial Network (GAN) based on Convolution Neural Network (CNN) is proposed to faces this challenge. The advantage of the GAN is represented by its architecture which consists of a generator and discriminator to detect real or fake image generated by the machine. In this paper, we adopt the advantage of GAN and combine with YOLO algorithm to identify similar music instruments. YOLO is fast Region based CNN with powerful computation. Using Deep Convolution YOLO-GAN will enhance the capability of YOLO detection process and outperform the original YOLO capability.