DNN Image Sensor Trigger Control for Power Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In image sensors equipped with a Deep Neural Network (DNN) processing unit, there is a risk of high power consumption when executing DNN processes at high-speed frame rates over extended periods.
Innovation Solution
An image capturing apparatus is provided with a DNN processing unit, a trigger unit that outputs a trigger signal to control the DNN process, and a control unit that manages the DNN process based on the trigger signal, allowing for the stopping or driving of the DNN process and its output to a subsequent stage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the DNN process is executed at a high-speed frame rate over a long amount of time, then the processing capability and recognition speed are improved, but the power consumption becomes large
Solution Approach 1:
The patent implements dynamic control of the DNN processing unit by switching between operation and stop states based on real-time trigger signals. The control unit dynamically adjusts the processing state of the DNN unit, allowing it to operate at high frame rates only when necessary (when trigger signals indicate motion or events) and stop during periods without events, thereby resolving the contradiction between maintaining high productivity and reducing power consumption.
Solution Approach 2:
The system employs periodic trigger signals to control the DNN processing unit's operation. Instead of continuous operation, the DNN unit is activated periodically based on incoming trigger signals from motion detection or event-based sensors. This periodic activation pattern allows the system to maintain high processing capability when needed while significantly reducing average power consumption during idle periods between triggers.
2Use of energy by moving object
If the DNN process is stopped, then the power consumption is reduced, but the processing capability and recognition speed are lowered
Solution Approach 1:
The control unit dynamically switches the DNN processing unit between stop and operation states based on real-time trigger signals. When triggers indicate motion or events requiring analysis, the system transitions to the operation state, instantly restoring full processing capability. This dynamic switching allows the system to maintain low power consumption during idle periods while ensuring high processing capability is immediately available when needed.
Solution Approach 2:
The system uses external trigger signals (from motion detectors, event-based sensors, or other input devices) to automatically control the DNN processing unit's operation state without requiring continuous external intervention. The trigger unit and control unit work together to enable the system to self-regulate its processing state based on environmental conditions, reducing power consumption while maintaining processing capability on-demand.
Data Source
AI summary
The present technique pertains to an image capturing apparatus and a signal processing method that enable power consumption to be lowered for a sensor equipped with a DNN (Deep Neural Network). The image capturing apparatus performs a DNN (Deep Neural Network) process on image data that is generated by image capturing, outputs a trigger signal for controlling stopping or driving of the DNN process, and, on the basis of the trigger signal, controls stopping of the DNN process or driving of the DNN process and output to a subsequent stage. The present technique can be applied to an image capturing apparatus that is provided with an image sensor which is equipped with a DNN.