Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Video bit enhancement method based on attention mechanism

A technology of attention and video, applied in image enhancement, television, image analysis and other directions, can solve the problem of flicker between frames of high-bit video sequences, and achieve the effect of improving the perceived visual quality

Active Publication Date: 2021-02-19
TIANJIN UNIV
View PDF2 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004]The above bit enhancement methods are all image-oriented, if they are applied to video sequences with low bit depth, the redundant information of the front and rear frames of the video cannot be obtained well The use of high-bit video sequences, and the resulting high-bit video sequences will appear inter-frame flicker and other phenomena

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Video bit enhancement method based on attention mechanism
  • Video bit enhancement method based on attention mechanism
  • Video bit enhancement method based on attention mechanism

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0052] Embodiments of the present invention include the following steps:

[0053] 101: From 16-bit Sintel database [9] Randomly select the original 1000 groups of video sequences, each group of 5 video frames, and quantize it to 4-bit depth, and apply the zero-filling algorithm to the 4-bit depth video sequence to expand the 4-bit video sequence into a 16-bit depth video sequence , the 16-bit-depth video frame extended by applying the zero-fill algorithm is called a rough high-bit-depth video frame;

[0054] 102: In this embodiment, the codec is used as the basic network architecture, and a global attention alignment module is added to the head of the encoder. This module can capture long-distance dependencies by calculating the correlation between intra-frame and inter-frame video sequences, and perform implicit Motion Estimation and Motion Compensation (ME&MC); add a target-guided semantic attention module at the connection between the encoder and decoder, which fuses the s...

Embodiment 2

[0059] Below in conjunction with specific experimental data, the effect evaluation of the embodiment 1 scheme is carried out, see the following description for details:

[0060] 301: Data composition

[0061] The test set consists of 50 groups of 16-bit continuous video frames randomly selected from the Sintel database that do not repeat the training set and 30 groups of 16-bit continuous video frames randomly selected from the TOS database, each group contains 5 frames of images.

[0062] 302: Evaluation Criteria

[0063] The present invention mainly adopts two kinds of evaluation indicators to evaluate the quality of the reconstructed high bit depth video frame:

[0064] Peak Signal to Noise Ratio (PSNR) is a commonly used objective image quality assessment method for evaluating images.

[0065] Structural Similarity Index (SSIM) [12] It is an index to measure the structural similarity of two images. This index measures the similarity of two images from the perspectives ...

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

No PUM Login to View More

Abstract

The invention discloses a video bit enhancement method based on an attention mechanism. The method comprises the following steps: establishing a video bit enhancement model based on the attention mechanism; randomly selecting a set number of original video sequence groups with high bit depth from an image enhancement database to construct a training data set; training a video bit enhancement modelbased on an attention mechanism by using the constructed training data set; selecting a video sequence group from an image enhancement database to form a test set, and testing the trained attention mechanism-based video bit enhancement model; applying a zero filling algorithm to video signals needing to be enhanced to obtain high-bit-depth video signals, sequentially inputting the high-bit-depthvideo signals into the tested video bit enhancement model based on the attention mechanism by taking five frames as a group, and adding output results of the high-bit-depth video signals to middle frames of the correspondingly input video sequence groups so as to sequentially and correspondingly obtain enhanced middle frames. According to the invention, the semantic attention matrix related to thetarget feature map is generated on the feature level, so that the perception visual quality is improved.

Description

technical field [0001] The invention relates to a video bit enhancement method. In particular, it involves an attention-based video bit enhancement method. Background technique [0002] Multimedia resources such as images and videos carry a wealth of information, and people can quickly understand what is happening in the outside world through images and videos. Since the birth of video recording devices and display devices, people have been working on how to obtain and display higher quality images and videos. In order to pursue a better visual experience, people have proposed a high dynamic range (High Dynamic Range, HDR) technology, which uses a higher dynamic range and more bit depth (usually 10 or 12 bits) to represent a pixel. Images and videos with high dynamic range can show richer colors, finer color transitions, and more realistic texture details. At present, with the development of technology, ultra-high-definition displays and HDR displays are becoming a common...

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
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/55H04N17/00
CPCG06T7/55H04N17/00G06T2207/20172G06T2207/10016
Inventor 刘婧杨紫雯于洁潇
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com