Adaptive Video Post-Processing for Compression Artifact Mitigation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Video codecs introduce noticeable compression artifacts during encoding and decoding, and conferencing tools often suffer from quality degradation due to network congestion or bandwidth limitations, leading to suboptimal video quality.

Innovation Solution

Implementing adaptive post-processing techniques in conferencing tools using machine learning models that selectively apply post-processing operations based on scenario detection, segmentation, and video quality analysis to enhance decoded video quality without significantly increasing network bandwidth.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If lossy compression is used to reduce bandwidth consumption, then network bandwidth usage is reduced, but video quality deteriorates due to compression artifacts

Engineering Contradiction:

Improvenetwork bandwidth consumptionVSAvoidvideo quality

Core Design Contradiction:

Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent applies machine learning post-processing models to convert the harmful compression artifacts into beneficial quality improvements. The models analyze decoded video frames and reconstruct lost high-frequency details, effectively transforming the degraded compressed video back into high-quality video without requiring additional bandwidth

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

Solution Approach 2:

The patent introduces machine learning post-processing models as an intermediary between the video decoder and the display. These models act as a mediator that takes the compressed decoded video as input and produces high-quality output by adding back the frequency information lost during compression, without affecting the original transmission

Inventive Principle:

Principle #24Intermediary (Mediator)

2Manufacturing precision

If video is encoded at high quality to maintain video quality, then video quality is improved, but network bandwidth consumption increases

Engineering Contradiction:

Improvevideo qualityVSAvoidnetwork bandwidth consumption

Core Design Contradiction:

Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent uses machine learning models to create a high-quality copy of the video content from the low-quality compressed version. The post-processing models generate synthetic high-frequency details that replicate what would have been present in the original high-quality video, effectively creating a quality copy without transmitting the actual high-bitrate data

Inventive Principle:

Principle #26Copying

3Productivity

If aggressive lossy compression is applied to save bandwidth, then bandwidth efficiency is improved, but compression artifacts increase

Engineering Contradiction:

Improvebandwidth efficiencyVSAvoidcompression artifacts

Core Design Contradiction:

ProductivityVSObject-generated harmful factors

Solution Approach 1:

The patent directly addresses this contradiction by using machine learning post-processing to convert the harmful compression artifacts into beneficial quality improvements. The models are trained to recognize and reconstruct the specific types of artifacts generated by aggressive compression, transforming them back into natural-looking video content

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

Solution Approach 2:

The patent changes the parameter of video quality from the compression domain to the post-processing domain. Instead of maintaining quality during compression, the system accepts compressed quality and applies quality enhancement through machine learning models that adjust the frequency content and visual characteristics of the decoded video

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20260006156A1Machine learning models for adaptive post-processing using results of video quality analysis in conferencing tools

Publication Date: 2026.01.01 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Innovations in machine learning (“ML”) models used in adaptive post-processing of decoded video in a conferencing tool are described. For example, as part of post-processing of decoded video, a super-resolution/video restoration model increases spatial resolution (e.g., by interpolation between sample values), mitigates compression artifacts, and mitigates upscaling artifacts introduced when increasing spatial resolution. Or, as another example, as part of post-processing of decoded video, a video restoration model mitigates compression artifacts, without increasing spatial resolution. For adaptive post-processing, a post-processing model can be selectively applied depending on results of scenario detection, results of segmentation, and/or results of video quality analysis. With the innovations, a conferencing tool can in effect provide video at higher quality without significantly increasing the network bandwidth consumed by the video or, alternatively, provide video using less network bandwidth without significantly hurting the quality of the video.