The design of error-correcting output codes based deep forest for the micro-expression recognition.

A micro-expression (ME) is hard to fake and reflects people's true feelings, micro-expression recognition (MER) is useful in a wide range of applications, such as medical research, criminal detection, and national security. Nevertheless, the existing deep learning methods based on deep neural networks (DNNs) fail to handle it, owing to the limited amount of training data and the black-box structure. In order to cope with these issues, this paper proposes a novel deep forest (DF) based on error-correcting output codes (ECOC), named EDF. The ECOC is innovatively deployed to encourage diversity and help to summarize differences among classes from multiple perspectives. Compared to DNN-based models, EDF is a more general solution for the MER since it is not sensitive to the size of data sets and needn't other approaches such as transfer learning. On the other hand, EDF copies with the interpretability-accuracy trade-off to a certain extent. The experimental results on several data sets confirm that EDF outperforms most state-of-the-art methods. Furthermore, EDF is an interpretable method. It can reconstruct feature values by tracing and fusing the decision paths, from which we can easily determine important features. According to the results of the back-reconstruction process, the difference among classes can be summarized. [ABSTRACT FROM AUTHOR]