Video description data processing method, device and storage medium

Abstract

The invention provides a video description data processing method, a device and a storage medium, and the method comprises the steps: importing a video sequence, and segmenting the video sequence into a plurality of video pictures; performing feature segmentation analysis on all the video pictures through a preset convolutional neural network to obtain a plurality of shot data sets; performing merging analysis on all the lens data sets through the preset convolutional neural network to obtain a plurality of merged lens data sets; performing feature extraction on the plurality of merged shot data sets through the preset convolutional neural network to obtain a video description feature sequence; and converting the video description feature sequence into video description information through a preset video description model. The method does not need to generate the final description after the text description is generated for each shot data, directly converts a natural language problem into an image problem, reduces the redundancy of the generated description, and improves the fluency of the text description.

Description

technical field [0001] The present invention mainly relates to the technical field of video processing, in particular to a video description data processing method, device and storage medium. Background technique [0002] At present, video description mainly includes machine translation accuracy indicators, sentence fluency indicators, etc., but how to solve the video fluency is a thorny problem. The existing technology is to generate several shot data sets according to the video shot segmentation, and then input each shot data into the convolutional neural network to generate a series of features, and then input these features into the video description model to generate sentences. In this process, it will cause complex calculations. At the same time, if two shots with high similarity are input respectively, each time a shot is input into the convolutional neural network, many similar features will be generated. In the video description model Describing each feature makes ...

AI Technical Summary

Problems solved by technology

In this process, it will cause complex calculations. At the same time, if two shots with high similarity are input respectively, each time a shot is input into the convolutional neural network, many similar features will be generated. In the video description model Describing each feature makes the calculation of the model very large, and the final description is very long-winded and not smooth, which is quite different from the manual description.

Images