Research on sketch instruction recognition technology ased on convolutional neural network

In order to solve the problems of low accuracy of traditional sketch instruction recognition, a sketch instruction recognition technology based on convolutional neural network is proposed. By constructing and optimizing the convolutional neural network model, a large number of sketch instruction samples are used for training, and the accuracy of the validation set is closely monitored throughout the training process, and the learning rate is dynamically adjusted in real time and based on this. With L2 regularization and Dropout dual insurance strategies, overfitting is synergistically suppressed. L2 regularization constraint weight scale and Dropout randomly inactivated neurons complement each other, which can improve the accuracy of the model in recognizing different sketch instructions and improve the human-computer interaction experience.