Neural network model training method and image retrieval method

The invention discloses a neural network model training method and an image retrieval method, and relates to the technical field of image processing. The training method comprises the following steps: selecting a positive example image and a negative example image of a query image based on a semi-hard negative example strategy to form a triple; inputting the image of the triple into a neural network model, performing feature extraction of different deep convolutional layers on the image, and extracting local features of a region of interest from an obtained feature map in combination with an attention network; aggregating the obtained local features to obtain global features; and comparing a loss function according to the global features, and updating the weight of the neural network model through a back propagation algorithm until a preset training stop condition is reached, thereby completing training. The training complexity can be effectively reduced, and the training speed is increased. 本发明公开了神经网络模型的训练方法