Smart Respiratory Sound Classification for Pulmonary Infiltrates Pre-Diagnosis
Coronavirus-related disease diagnosis system should be accurate and efficient. To improve the traditional diagnosis system, this project proposes not only a novel framework of deep neural networks-aid medical diagnosis system but also a large-scale respiratory sound data and label collection, Deep-SoundNet (DSN), to train the proposed system. Since the DSN contains three different types of labels, including the name of the disease, keywords, and clinical description, they can be exploited to train the system for automatic speech recognition (ASR). The main goal of this work is to serve as a benchmark to help the community in building a more accurate and efficient deep-learning-based automatic diagnosis system of coronavirus.
Winning Team
| Huck Yang is a PhD student in electrical and computer engineering at Georgia Tech. He was the former president of the Georgia Tech Taiwanese student association. Huck received his BS degree of electrical engineering from National Taiwan University in 2016. He was an Applied Scientist Intern on Amazon Alexa team in 2020. His interests includes education, modern art, and community service. |
Yi-Chieh Liu was a graduate student in the Department of Computational Science and Engineering at Georgia Tech. He received his BS degree from National Taiwan University in 2018. He was an Applied Scientist Intern on Amazon Alexa team. |
Comments