Scene Change Detection Circuitry for Compute Device Power Management

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Compute devices face challenges in balancing power management and user experience, as existing approaches either waste power by remaining always on or require slow boot times when waking up, and human presence sensors like camera-based systems are costly and inefficient.

Innovation Solution

Implementing a system that detects scene changes using low-resolution images to transition between sleep and wake modes efficiently, reducing power consumption and boot times by using hardware accelerator circuitry to process low-resolution statistic images and adjust computation conditions accordingly.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If the compute device remains always on to provide immediate responsiveness, then user experience is improved, but power consumption increases

Engineering Contradiction:
ImproveresponsivenessVSAvoidpower consumption
Core Design Contradiction:
Ease of operationVSUse of energy by moving object

Solution Approach 1:

The system dynamically transitions between sleep mode and wake mode based on detected scene changes. The compute device adjusts its operational state in real-time, being fully responsive when activity is detected and conserving power when idle, thus resolving the contradiction between constant responsiveness and power consumption.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The low-resolution image sensor continuously monitors for scene changes in advance, detecting user presence before full system activation is needed. This preliminary detection allows the device to wake up proactively rather than reactively, improving perceived responsiveness while maintaining power savings during idle periods.

Inventive Principle:
Principle #10Preliminary action

2Ease of operation

If the compute device wakes up from sleep mode, then responsiveness is improved, but boot time increases

Engineering Contradiction:
ImproveresponsivenessVSAvoidboot time
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The system segments the wake-up process into two stages: first, the low-resolution image sensor detects scene changes and triggers a partial wake-up; second, the main processor activates only when needed. This segmentation allows the device to prepare for responsiveness without requiring a full system boot, significantly reducing boot time while maintaining user experience.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Instead of performing a complete system boot upon every wake trigger, the device performs a partial wake-up that activates only the necessary components to maintain responsiveness. The full processing power is activated only when actually needed, reducing the time penalty associated with waking from sleep mode.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If traditional camera-based human presence sensors are used, then detection accuracy is improved, but device cost increases

Engineering Contradiction:
Improvedetection accuracyVSAvoiddevice cost
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces expensive, complex camera-based presence sensors with a low-resolution image sensor that is significantly cheaper and simpler. While the low-resolution sensor has limitations, it provides sufficient detection accuracy for the specific use case of detecting scene changes and user presence, thereby reducing device cost without sacrificing essential functionality.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

Solution Approach 2:

The system changes the resolution parameter of the image sensor from high (traditional camera) to low, which dramatically reduces cost and complexity. The low-resolution sensor is optimized specifically for detecting scene changes rather than capturing detailed images, making it a cost-effective solution that maintains adequate detection accuracy for presence sensing applications.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS12032972B2Methods, apparatus, and articles of manufacture to improve performance of a compute device by detecting a scene change
Publication Date: 2024.07.09 INTEL CORP
  • US12032972B2 patent drawing
  • US12032972B2 patent drawing
  • US12032972B2 patent drawing

AI summary

Methods, apparatus, systems, and articles of manufacture are disclosed to improve performance of a compute device by detecting a scene change. An example apparatus includes scene change detection circuitry and interrupt circuitry. The example scene change detection circuitry is to determine a first score value for a first metric of similarity between a first image of a field of view (FOV) of an image sensor and a second image of the FOV, determine a second score value for a second metric of similarity between the first image and the second image, and compute a composite score value based on the first score value and the second score value. The example interrupt circuitry is to generate an interrupt to processor circuitry of the compute device to cause the processor circuitry to adjust a computation condition of the compute device based on the composite score.