Micro-expression recognition method based on long-short-term memory network

The invention provides a micro-expression recognition method based on a long-term and short-term memory network. The micro-expression recognition method comprises the following steps: A, micro-expression sample processing: 1) processing sample data and label data of a micro-expression data set; 2) constructing a hierarchical data format hdf5 training and testing file; B, long-short-term memory network training: 3) constructing a convolutional neural network model, finely adjusting parameters for the constructed convolutional neural network model, and training a flexible maximum loss function to optimize the parameters; 4) fixing parameters of the trained convolutional neural network model, and training a long-short-term memory network model to minimize a loss function of a specific dimension; and C, micro-expression recognition: initializing test parameters according to the trained long-short-term memory network model, sending a sample for testing into the trained long-short-term memory network model, and out