Spectral unmixing is one of the most important quantitative analysis tasks in hyperspectral data processing. Conventional physics-based models are characterized by clear interpretation. However, due to the complex mixture mechanism and limited nonlinearity modeling capacity, these models may not be accurate, especially, in analyzing scenes with unknown physical characteristics. Data-driven methods have developed rapidly in recent years, in particular deep learning methods as they possess superior capability in modeling complex and nonlinear systems. Simply transferring these methods as black-boxes to conduct unmixing may lead to low physical interpretability and generalization ability. Consequently, several contributions have been dedicated to integrating advantages of both physics-based models and data-driven methods. In this article, we present an overview of recent advances on this topic from several aspects, including deep neural network (DNN) structures design, prior capturing and loss design, and summarise these methods in a common mathematical optimization framework. In addition, relevant remarks and discussions are conducted made for providing further understanding and prospective improvement of the methods. The related source codes and data are collected and made available at http://github.com/xiuheng-wang/awesome-hyperspectral-image-unmixing.