Dynamic Image Similarity Threshold Adjustment

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional image processing methods struggle to dynamically adjust the similarity threshold for displaying similar images, leading to suboptimal grouping and presentation of images based on varying user preferences.

Innovation Solution

A method and device that obtain image sets based on a similarity threshold, identify operation instructions, and update the threshold accordingly to adapt to user interactions, such as extracting or adding images, thereby adjusting the similarity threshold to generate image sets that align with different image similarity requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a fixed similarity threshold is used to group images, then the image grouping process is simple and fast, but the grouping accuracy does not adapt to varying user preferences

Engineering Contradiction:

Improvegrouping accuracyVSAvoidthreshold adjustment mechanism

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements dynamic adjustment of the similarity threshold based on user interactions with image sets. When users perform operations such as extracting images from a set or adding images to a set, the system automatically updates the threshold value to better reflect user preferences, transforming a static threshold into a dynamic adaptive parameter

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system uses user operations on image sets as feedback signals to adjust the similarity threshold. By monitoring whether users extract images from sets (indicating desire for stricter grouping) or add images to sets (indicating acceptance of looser grouping), the system continuously optimizes the threshold to improve grouping accuracy

Inventive Principle:

Principle #23Feedback

2Adaptability or versatility

If the similarity threshold is updated frequently to adapt to user preferences, then the personalization of image grouping improves, but the processing time and system response delay increase

Engineering Contradiction:

Improvepersonalization capabilityVSAvoidprocessing delay

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system updates the similarity threshold periodically based on accumulated user interactions rather than continuously. Each update occurs after collecting sufficient feedback from user operations, balancing the need for personalization with the need to minimize processing delays

Inventive Principle:

Principle #19Periodic action

Solution Approach 2:

The system pre-calculates and stores image feature vectors and similarity metrics before user interactions occur. This preliminary processing allows for faster threshold adjustments when user feedback is received, reducing the time penalty associated with personalization

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS9911173B2Method and device for displaying images

Publication Date: 2018.03.06 XIAOMI INC

AI summary

A method for displaying a plurality of images is provided. The method includes: obtaining one or more image sets based on the plurality of images, wherein a similarity degree between each pair of images in each image set is greater than a similarity threshold; identifying an operation instruction triggered on at least one of the image sets; if the operation instruction triggered on the at least one of the image sets satisfies a predetermined updating condition, updating the similarity threshold; and displaying the plurality of images based on the updated similarity threshold.