Video super-division model training method and device and video super-division processing method and device
A model training and video technology, applied in the computer field, can solve the problems of unsatisfactory video over-score results and the failure to make good use of video frame information, and achieve the effect of ensuring the effect of maintaining details.
Active Publication Date: 2022-03-01
BEIJING BAIDU NETCOM SCI & TECH CO LTD
View PDF6 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Some early video super-resolution methods continued the idea of image super-resolution, and did not make good use of continuous video frame information, resulting in unsatisfactory video super-resolution results
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
[0021] Exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and they should be regarded as exemplary only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
[0022] For video frames in the same video, the content source may be the same, but there is motion offset, scene offset, or light offset. In this case, how to align video frames and find a common reference point is a problem that needs to be solved for video frame alignment.
[0023] The video alignment method aligns adjacent frames in the video frame with the target f...
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 provides a video super-division model training method and device, and relates to the technical fields of computer vision, deep learning and the like. The specific implementation scheme is as follows: acquiring a video set comprising more than two video frames; a pre-established video super-division network is obtained, the video super-division network comprises a sliding window alignment subnet, a cyclic alignment subnet and a refined alignment subnet, the sliding window alignment subnet is used for carrying out feature extraction and fusion on a plurality of continuous video frames, the cyclic alignment subnet is used for carrying out bidirectional propagation on features output by a sliding window, and the refined alignment subnet is used for carrying out feature extraction and fusion on a plurality of continuous video frames; reconfiguration features and alignment parameters are generated, the refined alignment subnet carries out realignment and bidirectional propagation on the reconfiguration features based on the alignment parameters, and super-divided video frames are obtained; the method comprises the following training steps: inputting continuous video frames selected from a video set into a video super-division network, and calculating a loss value of the video super-division network; and obtaining a video super-division model based on the loss value of the video super-division network. According to the embodiment, the video super-division effect of the model is improved.
Description
technical field [0001] The present disclosure relates to the field of computer technology, specifically to computer vision, deep learning and other technical fields, and in particular to a video super-resolution model training method and device, a video super-resolution processing method and device, electronic equipment, computer-readable storage media, and computer programs product. Background technique [0002] The Video Super-Resolution task is to recover the corresponding high-resolution counterpart from a given low-resolution video. In recent years, with the explosive growth of Internet video data, video super-resolution technology has received great attention from researchers. Similar to image super-resolution, video super-resolution is also an ill-posed problem. However, unlike image super-resolution, video super-resolution not only needs to pay attention to the corresponding low-resolution frames, but also needs to utilize the information of consecutive frames in t...
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 IPC IPC(8): G06T3/40G06N3/04G06N3/08
CPCG06T3/4053G06T3/4046G06N3/084
Inventor 王娜江列霖党青青赖宝华
Owner BEIJING BAIDU NETCOM SCI & TECH CO LTD