Spatial Resampling Filters for Bitrate-Efficient Video Decoding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing video coding and decoding systems face challenges in efficiently managing bitrate requirements and adapting to different video consumption scenarios, particularly for machine vision tasks where video quality and resolution needs differ between human and machine consumption.

Innovation Solution

Implementing spatial downsampling and resampling techniques in video coding and decoding systems, using spatial resampling flags and filters to adapt video data based on the consumption scenario, whether for human or machine use.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If spatial downsampling is applied to reduce data volume, then transmission and storage efficiency is improved, but video quality and resolution are degraded

Engineering Contradiction:

Improvedata volumeVSAvoidvideo quality

Core Design Contradiction:

Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent applies different quality levels to different regions and scenarios: full resolution for human vision tasks, downsampled resolution for machine vision tasks, and selectively processed regions for object detection. This allows the system to optimize data volume while preserving quality where needed.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent dynamically adjusts resolution and downsampling parameters based on the specific application scenario, task requirements, and content characteristics. The system can switch between different resolution levels and apply adaptive downsampling filters to maintain optimal quality for each scenario.

Inventive Principle:

Principle #15Dynamics

2Adaptability or versatility

If different resolution requirements are met for human and machine vision tasks, then adaptability is improved, but system complexity increases

Engineering Contradiction:

Improveadaptability to different consumption scenariosVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the video processing pipeline into distinct paths: one for human vision tasks maintaining full resolution, and another for machine vision tasks using downsampled resolution. This segmentation allows independent optimization for each task type without increasing overall system complexity.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent creates a universal video coding system that can serve multiple functions: human vision consumption, machine vision tasks, object detection, and tracking. The same encoded bitstream can be decoded and processed differently based on the application scenario, reducing the need for separate processing systems.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS20260025498A1Spatial resampling in video coding and decoding systems

Publication Date: 2026.01.22 TENCENT AMERICA LLC

AI summary

This disclosure relates generally to video coding/decoding and particularly for spatial downsampling and/or resampling in video coding and/or decoding systems. One method includes obtaining, by a device, a coded video bitstream; determining, by the device from the coded video bitstream, a spatial resampling flag for a picture frame; and when the spatial resampling flag indicates that spatial resampling is enabled for the picture frame: determining, by the device from the coded video bitstream, an index indicating a spatial resampling filter, and decoding, by the device, the coded video bitstream by generating spatial resampling data based on the spatial resampling filter.