DNN Popularity Model for Image Content Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for evaluating image popularity on social networks inaccurately combine visual and non-visual factors, leading to inefficient allocation of computational and storage resources.
Innovation Solution
A deep neural network (DNN) based intrinsic popularity assessment model is used to determine an intrinsic popularity score for images, emphasizing visual content and allowing for more accurate popularity evaluation and efficient resource management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods combine visual and non-visual factors with equal importance for popularity evaluation, then the evaluation process is simple, but the accuracy of popularity prediction deteriorates
Solution Approach 1:
The patent segments the popularity evaluation into two distinct components: visual factors (image content, composition, aesthetics) and non-visual factors (user statistics, upload time, caption). By separating these factors and evaluating them independently with appropriate weighting, the model achieves more accurate popularity prediction while maintaining manageable complexity through modular architecture.
Solution Approach 2:
The patent introduces intrinsic popularity scores as a new parameter that transforms the evaluation from treating all factors equally to assigning differentiated weights based on their actual contribution to popularity. This parameter change enables the system to capture the nuanced relationships between visual and non-visual factors, improving prediction accuracy without proportionally increasing complexity.
2Productivity
If computational and storage resources are allocated uniformly to all content, then resource management is simple, but resource utilization efficiency deteriorates
Solution Approach 1:
The patent applies local quality by allocating computational and storage resources differently based on the intrinsic popularity score of each piece of content. High-popularity content receives more resources for processing and storage, while low-popularity content receives fewer resources. This differentiated allocation optimizes resource utilization efficiency by matching resource investment with actual content value and user engagement potential.
3Quantity of substance
If variable compression rates are applied based on intrinsic popularity scores, then storage efficiency is improved, but system complexity increases
Solution Approach 1:
The patent implements dynamic compression where the compression rate is not fixed but varies based on the intrinsic popularity score of each content item. The system dynamically adjusts compression parameters to balance storage efficiency with quality preservation, applying higher compression to low-popularity content and lower compression to high-popularity content. This dynamic approach optimizes overall storage utilization while adapting to the specific characteristics of each content piece.
Data Source
AI summary
The present application provides methods, devices and computer readable media for intrinsic popularity evaluation and content compression based thereon. In an embodiment, there is provided a method of intrinsic popularity evaluation. The method comprises: receiving an image from a social network; and determining an intrinsic popularity score for the image using a deep neural network (DNN) based intrinsic popularity assessment model.