Face Stylization Algorithm Switching for Varied Image Displays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies lack diversity and interactivity in displaying style images, leading to monotonous user experiences.
Innovation Solution
A method and apparatus for generating style images by randomly selecting a face stylization algorithm, performing stylization processing on face image regions, and switching displays using preset rendering parameters, incorporating interactive and random effects to enhance user engagement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a fixed face stylization algorithm is used for processing face image regions, then the processing efficiency is maintained, but the diversity and interactivity of style image generation is reduced
Solution Approach 1:
The patent applies dynamics by making the face stylization algorithm selection dynamic rather than fixed. The system randomly selects different stylization algorithms from a preset collection based on rendering parameters, enabling the processing mechanism to adapt and change according to different display scenarios and user preferences, thereby improving diversity without requiring complex manual configuration
Solution Approach 2:
The patent utilizes parameter changes by varying the rendering parameters that control which stylization algorithm is selected and how the transition is performed. By adjusting parameters such as transition duration, opacity, and algorithm selection seeds, the system achieves diverse stylistic effects and interactive control without fundamentally changing the core processing architecture
2Speed
If direct stylization processing is performed on face images, then the processing speed is maintained, but the smoothness of transition between original and stylized images is reduced
Solution Approach 1:
The patent applies preliminary action by pre-processing the stylized image in the background before it needs to be displayed. The system performs the computationally intensive stylization processing on the face image region in advance, prepares the transition parameters, and queues the rendering operations, so that when the display is triggered, the transition can occur smoothly without noticeable delays or abrupt changes
Solution Approach 2:
The patent introduces an intermediary rendering mechanism that mediates between the original image and the fully stylized image. Instead of directly switching between the two extremes, the system uses intermediate transition frames with controlled opacity changes and blending effects, creating a smooth visual transition that maintains both processing efficiency and visual comfort
3Adaptability or versatility
If multiple face stylization algorithms are provided for selection, then the diversity of style images is improved, but the operation complexity for users increases
Solution Approach 1:
The patent applies self-service by enabling the system to automatically select and apply appropriate face stylization algorithms based on preset criteria and rendering parameters, without requiring users to manually choose from multiple options. The system autonomously manages the diversity of effects by randomly or contextually selecting from the preset collection, thereby providing variety while maintaining simple, intuitive user interaction
Solution Approach 2:
The patent utilizes periodic action by implementing a random selection mechanism that cycles through or samples different stylization algorithms from the preset collection. This periodic sampling approach ensures that users experience a variety of styles over time through repeated interactions, maintaining ease of operation while delivering diverse stylistic results
Data Source
AI summary
The embodiments of the present disclosure relate to an style image generation method and apparatus, a device and a medium, wherein the method comprises: acquiring an image to be processed comprising a face image region, acquiring a target face stylization algorithm from preset face stylization algorithms, and performing a stylization processing on the face image region based on the target face stylization algorithm to obtain a face stylization image; performing a processing based on the face stylization image and the image to be processed to obtain a target image; and switching from displaying the image to be processed to displaying the target image according to a preset rendering parameter.