Image Artifact Modification via Adaptive Neural Network Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems in electronic devices, such as smartphones, face user experience issues due to the time-consuming process of removing complex image artifacts like reflections using large neural networks, which delays the viewing of corrected images.
Innovation Solution
The implementation of a lightweight neural network system that activates pre-trained neural networks or neural network layers based on user input parameters like speed, length, or pressure, allowing for efficient and real-time artifact modification without the need for extensive processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a large neural network is used to remove complex image artifacts, then artifact removal effectiveness is improved, but processing time increases
Solution Approach 1:
The patent segments the artifact removal process into multiple stages: first using a lightweight neural network for quick initial processing, then selectively applying a large neural network only to specific regions containing complex artifacts that require more sophisticated processing. This segmentation allows the system to achieve high artifact removal effectiveness while minimizing overall processing time by avoiding full application of the computationally intensive large network.
Solution Approach 2:
The patent implements partial action by applying the large neural network only to specific regions of interest rather than the entire image. The system identifies regions containing complex artifacts and applies the large network selectively to those areas, achieving effective artifact removal where needed while reducing total processing time and computational resources required.
2Manufacturing precision
If a large neural network is used to process high-resolution images, then artifact removal quality is improved, but user experience deteriorates due to waiting time
Solution Approach 1:
The patent applies preliminary action by first processing the image with a lightweight neural network to remove common and simple artifacts quickly. This preliminary processing provides immediate visible results to the user, improving user experience by reducing waiting time. Subsequently, the system identifies and processes only the remaining complex artifacts using the large neural network, ensuring high removal quality without requiring the user to wait for complete processing of the entire image.
3Reliability
If extensive neural network processing is applied, then artifact correction completeness is improved, but processing efficiency decreases
Solution Approach 1:
The patent implements dynamics by creating an adaptive, multi-stage processing system that dynamically adjusts the level of processing applied to different regions of the image. The system first applies lightweight processing globally, then dynamically identifies regions requiring more intensive processing and applies the large neural network selectively. This dynamic approach ensures complete artifact correction while maintaining high processing efficiency by avoiding unnecessary application of computationally intensive processing to already-clean regions.
Data Source
AI summary
A method 300B includes detecting a user input indicative of a trigger to modify the artifact of the image displayed at a user interface of an electronic device 100. Furthermore, the method 300B includes determining an artifact modification parameter based on a characteristic of the user input. Furthermore, the method 300B includes modifying the artifact in the image based on the artifact modification parameter.