Lifelong Deep Learning

This chapter introduces an important future development direction of embedded AI, lifelong deep learning. By analyzing the shortcomings and causes of traditional deep learning, we clarify the goals and characteristics of lifelong deep learning and explore some methods to implement lifelong deep neural networks, such as dual learning systems, real-time updates, memory merging, and adaptation to real scenarios. Finally, the advantages brought by the combination of lifelong deep neural network and embedded AI are summarized, such as autonomous learning, federated learning, etc.