Digital Image Professionalism Scoring and Cropping via DCNN

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The challenge lies in automatically enhancing the perceived professionalism of digital images, particularly in professional contexts, where traditional quality metrics fail to capture nuances such as attire and setting appropriateness, which vary by industry and location.

Innovation Solution

A Deep Convolutional Neural Network (DCNN) is employed to generate professionalism scores for digital images by identifying relevant features, including cropping, color balance, and depth of field, and then applies transformations like cropping and rotation to improve the image's professionalism score.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional image quality metrics are used to evaluate digital images, then technical aspects like lighting and framing can be quantified, but professionalism nuances such as attire and setting appropriateness cannot be captured

Engineering Contradiction:
Improveimage quality measurementVSAvoidprofessionalism assessment
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent transforms the evaluation from traditional technical parameters (lighting, framing) to include professionalism parameters (attire, setting, pose) by using a DCNN model that learns to assess multiple dimensions of image quality, enabling comprehensive evaluation that adapts to professional context requirements

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent replaces manual or rule-based image quality assessment with an automated DCNN-based system that can independently evaluate both technical and professionalism aspects, eliminating the need for human reviewers and enabling scalable, consistent evaluation across large numbers of images

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If manual image selection and editing is performed to ensure professionalism, then image quality can be controlled, but time consumption increases significantly

Engineering Contradiction:
Improveimage professionalismVSAvoidimage processing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent enables images to evaluate and transform themselves automatically through the DCNN model, which identifies professionalism issues and applies appropriate transformations (cropping, rotation, color adjustment) without human intervention, making the system self-sufficient and eliminating manual processing time

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent performs image evaluation and transformation automatically during the image upload or profile creation process, so that by the time the image is displayed, it has already been optimized for professionalism, eliminating the need for subsequent manual editing and ensuring readiness before use

Inventive Principle:
Principle #10Preliminary action

3Productivity

If automated cropping and transformation is applied to improve professionalism, then image optimization is achieved, but the original image composition may be altered

Engineering Contradiction:
Improveimage optimization efficiencyVSAvoidimage composition
Core Design Contradiction:
ProductivityVSShape

Solution Approach 1:

The patent applies transformations locally and selectively based on the specific issues detected in each image region, such as cropping only the portions that contain inappropriate background elements or adjusting color balance only in areas with poor lighting, rather than uniformly transforming the entire image, thus preserving important compositional elements while improving professionalism

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS10043240B2Optimal cropping of digital image based on professionalism score of subject
Publication Date: 2018.08.07 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

In an example embodiment, an optimal cropping of a digital image is determined. A machine learning algorithm is used to generate a professionalism score for the digital image, the utilizing a machine learning algorithm comprising a training mode where a plurality of sample images with labeled professionalism scores are used to train a classification function in a model that produces as professionalism score as output; and an analysis mode where the model is used to generate a professionalism score for the digital image. Then, the professionalism score is used as an input to a discrete variable optimization algorithm to determine an optimum cropped version of the digital image from a plurality of possible cropped versions of the digital image using the classification function.