No-Reference Image Quality Assessment With Asynchronous Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image quality assessment methods are subjective, costly, and difficult to scale due to the reliance on human evaluation and full-reference models, which do not align well with human visual perception, and lack generalization across different aesthetics and quality measurements.
Innovation Solution
A no-reference framework using a deep neural network that automatically extracts visual features from image pixels, trained on human-scored datasets, to objectively assess image quality, including overall quality, exposure, blurriness, and color tone, with asynchronous model inference on user devices to avoid latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human evaluation and full-reference models are used for image quality assessment, then measurement precision is improved, but productivity is worsened due to being costly and difficult to scale
Solution Approach 1:
The patent replaces the mechanical system of human evaluation with an automated no-reference image quality assessment model that processes images algorithmically. This substitution eliminates the need for costly human reviewers while maintaining quality assessment capabilities through machine learning-based metrics that evaluate images without requiring reference images or human intervention.
Solution Approach 2:
The patent creates a computational model that replicates human visual perception characteristics through algorithms that mimic how humans perceive image quality. By copying the essential aspects of human evaluation into a digital model, the system achieves both automation and perceptual accuracy, enabling scalable assessment that reflects human judgment without requiring actual human reviewers.
2Measurement precision
If full-reference models are used for image quality assessment, then measurement precision is improved, but device complexity is worsened due to requiring reference images
Solution Approach 1:
The patent extracts and removes the requirement for reference images from the quality assessment system. By developing a no-reference model that evaluates images independently without needing original or reference images for comparison, the system simplifies the overall structure while maintaining assessment capabilities through intrinsic image analysis methods.
3Measurement precision
If synchronous image quality assessment is performed on user devices, then measurement precision is improved, but loss of time is worsened due to latency
Solution Approach 1:
The patent implements asynchronous image quality assessment that performs quality evaluation in the background without blocking the main application flow. By executing quality assessment as a separate, non-critical task that does not require immediate completion, the system maintains user interface responsiveness while still performing comprehensive quality analysis, accepting that the assessment completes after the user interaction has already occurred.
Data Source
AI summary
A method of image quality assessment, performed by one or more processors in an image capture device, is disclosed. The method comprising receiving notification of capture of an image and in response, initiating an image assessment task to assess the quality of the image. The assessment task comprises determining suitability of the image for image quality assessment, running an image quality assessment model on the image to generate image quality assessment results, collecting data related to the capture of the image, and transmitting the results to an image quality assessment repository. The image assessment task may be a lower priority asynchronous task.