Image Description Generation Model Based on Attention Mechanism

Image description generation refers to translating a picture into a description text. This paper proposes an image title generation model based on the attention mechanism based on the encoder-decoder architecture. The encoder part uses the perception V3 network based on a convolutional neural network to extract image features, and the decoder will use the GRU network based on a cyclic neural network to decode the image coding information and output the sequence; with the adoption of attention mechanism, the decoder can generate text vector through the hidden state caused by image coding information and the output sequence of the decoder, to improve the prediction effect of the decoder on words.