Research on Deep Learning-based Cross-disciplinary Applications

Currently, with the development of artificial intelligence, deep learning is widely applied in several disciplines. This paper presents several cross-disciplinary studies on deep learning, including deep learning for preserving ancient cultural heritage, solving social problems, and industry application. Ancient documents contain a great deal of historical information that needs to be preserved. Therefore, the deep learning-based You Only Look Once is applied to the preservation of ancient cursive script and Oracle Bone inscriptions. Moreover, deep learning effectively solves the problem of lack of labor. Deep learning-based robots are used for tableware recycling. The object detection algorithm provides fast recognition. At the same time, the application of deep learning-based image classification algorithms for egg and apple classification. However, deep learning models are becoming more complex. Finally, research on deep learning in high-performance computing with model compression and acceleration, FPGAs, and neural architecture search is discussed.