Publications with shared reference: Optimized cross-corpus speech emotion recognition framework based on normalized 1D convolutional neural network with data augmentation and feature selection