Game Frame Classification for Adaptive GPU Driver Rendering

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing video processing systems lack efficiency in adjusting rendering and presenting operations based on the current stage of video game applications, leading to higher power consumption and visual artifacts due to mismatches in settings and application stages.

Innovation Solution

Implementing an image classification data model, such as a trained neural network, to identify the category of a rendered video frame and adjust rendering and presenting operations accordingly, using machine learning techniques to optimize parameters like FPS, motion-compensated frame interpolation, and resource allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If video processing operations are performed at high settings regardless of application stage, then processing performance is maintained, but power consumption increases without performance benefit

Engineering Contradiction:

Improvevideo processing performanceVSAvoidpower consumption

Core Design Contradiction:

ProductivityVSUse of energy by moving object

Solution Approach 1:

The system dynamically adjusts rendering and presenting operations based on the detected stage of the video game application. The graphics driver classifies the current stage (menu, loading, active gameplay) and adapts processing settings accordingly, transitioning from high-performance modes during active gameplay to power-saving modes during static stages, thus resolving the contradiction between maintaining performance and reducing power consumption.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes processing parameters such as rendering quality, frame rate, and presentation settings based on the application stage classification. By modifying these parameters dynamically according to whether the application is in menu, loading, or active gameplay state, the system optimizes the balance between processing performance and power consumption.

Inventive Principle:

Principle #35Parameter changes

2Device complexity

If fixed rendering settings are used throughout the application, then device complexity is reduced, but visual artifacts occur due to stage-mismatched processing

Engineering Contradiction:

Improveprocessing control complexityVSAvoidvisual artifacts

Core Design Contradiction:

Device complexityVSObject-affected harmful factors

Solution Approach 1:

The system incorporates feedback from the rendered video frame itself to determine the current application stage. The graphics driver analyzes visual characteristics of the rendered frame to classify the stage, creating a closed-loop system that automatically adjusts processing settings to prevent visual artifacts without requiring complex external control mechanisms.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The rendering system performs self-classification of the current application stage by analyzing the rendered video frame properties. This self-service mechanism allows the system to autonomously determine when to apply different processing settings, eliminating the need for complex external control systems while preventing visual artifacts through stage-appropriate processing.

Inventive Principle:

Principle #25Self-service

3Speed

If real-time video processing is performed without stage awareness, then processing speed is maintained, but resource allocation becomes inefficient

Engineering Contradiction:

Improvevideo processing speedVSAvoidresource allocation efficiency

Core Design Contradiction:

SpeedVSProductivity

Solution Approach 1:

The system performs preliminary classification of the video frame to determine the current application stage before executing the rendering and presenting operations. This preliminary action enables the system to allocate resources more efficiently by preparing appropriate processing settings in advance based on the detected stage, maintaining real-time processing speed while improving resource allocation efficiency.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20250381481A1Game interface classification using ML and driver optimization

Publication Date: 2025.12.18 ATI TECHNOLOGIES ULC

AI summary

An apparatus and method for performing efficient video data processing. In various implementations, a computing system includes a client device executing a parallel data graphics application that processes multiple video frames. The application includes multiple iterations of a loop with each loop processing a single video frame such as rendering and presenting the rendered video frame to a display controller. The client device adjusts the rendering operation and the presenting operation for subsequent video frames based on an image type of the current video frame. The client device utilizes an image classification data model that relies on machine learning techniques to generate an indication specifying the category (image type) of multiple categories of the current video frame based on the rendered data of the video frame. Examples of the categories are a menu image, an application loading image, a scoreboard image, and an active gameplay image.