Video search method, storage medium, device and system based on deep learning
A deep learning and video search technology, applied in video data indexing, video data retrieval, video data query, etc., can solve problems such as few text descriptions and difficult to search live video, and achieve the effect of improving accuracy
Active Publication Date: 2021-09-07
WUHAN DOUYU NETWORK TECH CO LTD
View PDF6 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] Due to the various forms of live broadcast content, there are many differences between live video and ordinary video, graphics and texts. One of the biggest differences is that videos generated by the live broadcast industry have fewer text descriptions. There are few descriptions, which makes it difficult for users to search for relevant live videos when using the search function of the live broadcast website to search for live videos
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0027] The present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments.
[0028] see figure 1 As shown, the embodiment of the present invention provides a video search method based on deep learning, which is convenient for users to search and watch live videos. The video search method based on deep learning in the embodiment of the present invention specifically includes:
[0029] S1: Select a set number of samples in each video category of the live broadcast. The sample is a video frame, that is, a frame of video, and each sample is marked with the category to which the frame belongs and the frame contained in the frame. A short description of the content. The categories of live broadcast videos include outdoor, game, and singing and dancing. Outdoor videos are the images of the host performing outdoor live broadcasts. This type of video is mainly about people, outdoor buildings, and scenes. Game videos are the an...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a video search method, storage medium, device and system based on deep learning, and relates to the field of software engineering. The method includes selecting a set number of samples in each live video classification; inputting all samples into NeuralTalk2 The deep learning model is obtained after training and training; all live videos are input into the deep learning model, and the deep learning model outputs a description of the content contained in each input picture; all the content of each video output by the deep learning model is The description of the screen is counted, and at least one description with a high frequency of occurrence is used as the index of the current video; when the user searches, the index containing the keyword is matched according to the keyword entered by the user, and the video corresponding to the index is displayed as the search result . The invention can effectively improve the accuracy of users searching live video results.
Description
technical field [0001] The present invention relates to the field of software engineering, in particular to a video search method, storage medium, device and system based on deep learning. Background technique [0002] At present, the live broadcast industry is developing rapidly, and people can conveniently watch live broadcasts with various contents through smart mobile devices, which greatly enriches people's spare time life. [0003] Due to the various forms of live broadcast content, there are many differences between live video and ordinary video, graphics and texts. One of the biggest differences is that videos generated by the live broadcast industry have fewer text descriptions. There are few descriptions, which makes it difficult for users to search for relevant live videos when they use the search function of the live broadcast website to search for live videos. Contents of the invention [0004] In view of the defects existing in the prior art, the purpose of ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/732G06F16/783G06F16/71
CPCG06F16/71G06F16/73G06F16/783
Inventor 刘宇超张文明陈少杰
Owner WUHAN DOUYU NETWORK TECH CO LTD