Compressed Video ML Processing for Artifact-Free XR Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine learning models applied to compressed videos often generate inaccurate predictions due to artifacts introduced by video compression, leading to unrealistic XR presentations and resource wastage.
Innovation Solution
Train machine learning models to process compressed videos in a way that eliminates or minimizes artifacts, ensuring accurate predictions and realistic XR experiences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine learning models are applied to compressed videos, then processing speed and resource efficiency are improved, but prediction accuracy deteriorates due to compression artifacts
Solution Approach 1:
The patent applies deblocking filtering as a preliminary action before machine learning processing. The filter removes compression artifacts from the video data in advance, so that when the ML model processes the filtered data, it achieves both high processing speed (by working on compressed data) and high prediction accuracy (by eliminating artifact interference)
2Quantity of substance
If video compression is applied to reduce data size, then transmission and storage efficiency are improved, but image quality deteriorates due to introduced artifacts
Solution Approach 1:
The patent extracts and removes the harmful compression artifacts from the compressed video data through deblocking filtering. This allows the system to maintain the benefits of compressed data (small size for efficient transmission and storage) while eliminating the quality-degrading artifacts, thus preserving image quality without requiring uncompressed data
Data Source
AI summary
Methods and systems are disclosed for applying machine learning models to compressed videos. The system receives a video, depicting an object, that has previously been compressed using one or more video compression processes. The system analyzes, using one or more machine learning models, the video that has previously been compressed to generate a prediction corresponding to the object depicted in the video, with one or more artifacts resulting from application of the one or more machine learning models to the video that has been previously compressed being absent from the prediction. The system generates a visual output based on the prediction in which the one or more artifacts are absent.