Video classification method and model training method and device thereof, and electronic equipment

Abstract

The invention provides a video classification method, a model training method and device thereof, and electronic equipment. The training method comprises the following steps: extracting initial features of a plurality of video frames through a convolutional neural network; extracting final features of the plurality of video frames from the initial features through a recurrent neural network; inputting the final feature into an output network, and outputting a prediction result of the multi-frame video frame; determining a loss value of the prediction result through a preset loss prediction function; and training the initial model according to the loss value until parameters in the initial model converge to obtain a video classification model. According to the method, the convolutional neural network and the recurrent neural network are combined, so that the operand can be greatly reduced, and the model training and recognition efficiency is improved; and meanwhile, the association information between the video frames can be considered in the feature extraction process, so that the extracted features can accurately represent the video types, and the accuracy of video classificationis improved.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a video classification method and its model training method, device and electronic equipment. Background technique [0002] In related technologies, videos can be classified through a three-dimensional convolutional neural network, and the spatiotemporal features of the video can be extracted through three-dimensional convolution. However, the network parameters of the three-dimensional convolutional neural network are relatively large, resulting in high computational costs in the network training process and recognition process. , the time overhead is high; in addition, the layers of the three-dimensional convolutional neural network are shallow, and it is difficult to mine high-level semantic features, resulting in low video classification accuracy. Contents of the invention [0003] In view of this, the purpose of the present invention is to provide a video ...

AI Technical Summary

Problems solved by technology

[0002] In related technologies, videos can be classified through a three-dimensional convolutional neural network, and the spatiotemporal features of the video can be extracted through three-dimensional convolution. However, the network parameters of the three-dimensional convolutional neural network are relatively large, resulting in high computational costs in the network training process and recognition process. , the time overhead is large; in addition, the layers of the 3D convolutional neural network are shallow, and it is difficult to mine high-level semantic features, making the accuracy of video classification low

Images