Affective EEG-Based Person Identification Using Channel Attention Convolutional Neural Dense Connection Network

Figure 1

Overview of the proposed network architecture. There are three parts in the model: channel attention model, dense layer, and transition layer. The channel attention model is formed by a convolutional layer and a LeakyRelu activation function layer with one channel scales. The dense layer comprises of a convoluted and a concrete layer. The transition layer consists of a convolutional module and a max pooling layer which models can reduce the dimension of the concrete feature matrix.