Broadcast Display Gaze Control for Low-Power Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Intelligent terminals face high power consumption due to prolonged display usage, affecting system lifespan and efficiency.
Innovation Solution
A control method utilizing a pre-trained neural network model for gaze recognition, integrated with In-Memory Computing (IMC) technology, to determine display states based on user gaze, reducing power consumption by transitioning to sleep mode when no gaze is detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the display screen is kept on for continuous monitoring, then the monitoring function is maintained, but the power consumption increases and system lifespan is reduced
Solution Approach 1:
The display screen dynamically adjusts its state between awake and sleep modes based on real-time gaze detection. When no gaze is detected for a predetermined period, the system transitions to sleep mode to reduce power consumption. When gaze is detected, the system wakes up to restore monitoring function, thus adapting the display state to actual usage needs.
Solution Approach 2:
The system uses gaze recognition feedback to control display state transitions. The gaze detection module continuously monitors user attention and provides feedback to the control module, which adjusts the display screen state accordingly - waking it up when needed and putting it to sleep when not needed, creating a closed-loop control system.
2Loss of energy
If gaze recognition is performed continuously to control display states, then power saving is achieved, but recognition precision may be affected
Solution Approach 1:
The gaze recognition model is pre-trained using knowledge distillation before deployment. This preliminary training process transfers knowledge from a complex teacher model to a lightweight student model, ensuring that the model achieves high recognition precision even when running efficiently on embedded devices with limited computational resources.
Solution Approach 2:
The system adjusts the operational parameters of the gaze recognition model based on power consumption constraints. By using a knowledge-distilled lightweight model instead of a full-complexity model, the system maintains precision while operating within power limits, effectively changing the model parameter scale to match hardware capabilities.
Data Source
AI summary
The present disclosure provides a control method of a broadcast monitoring system, a control apparatus of a broadcast monitoring system, a computer device, and a computer storage medium, and belongs to the field of image recognition and terminal broadcast monitoring. The control method of a broadcast monitoring system includes: obtaining a detected image; performing gaze recognition on the detected image through a pre-trained target neural network model, to obtain a recognition result of the detected image; and sending the recognition result to a terminal, so that the terminal determines a display state based on at least the recognition result.