Mobile Image Mask Generation via Segmentation Engine
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Generating image masks for mobile devices is computationally intensive due to limited processing power and memory, making it challenging to apply image effects to specific regions of an image efficiently.
Innovation Solution
A system that uses a neural network-based segmentation engine to create image masks by resizing training images, implementing multi-scale training, and refining mask borders, allowing for efficient detection and labeling of image areas on mobile devices, enabling precise application of visual effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image mask generation methods are used on mobile devices, then accurate region labeling can be achieved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent segments the image processing task into distinct components: a segmentation engine that processes the input image to identify regions, and a separate machine learning model that classifies these regions. This division allows each component to be optimized independently, reducing overall computational complexity while maintaining accuracy.
Solution Approach 2:
The patent introduces an intermediary segmentation engine that acts as a bridge between the input image and the classification model. This intermediary prepares the image data by identifying and segmenting regions before they are fed to the classification model, making the subsequent classification more efficient and accurate while reducing the computational burden on mobile devices.
2Measurement precision
If high-precision image mask generation is implemented on mobile devices, then accurate image effects can be applied, but processing speed decreases due to limited computational resources
Solution Approach 1:
By segmenting the processing task into region identification and classification stages, the patent enables faster processing. The segmentation engine quickly identifies regions of interest, and the classification model efficiently processes these pre-segmented regions, achieving both speed and precision on mobile devices.
Solution Approach 2:
The segmentation engine performs preliminary action by pre-processing the image to identify and segment regions before classification. This preliminary segmentation reduces the complexity of the subsequent classification task, enabling faster processing while maintaining high precision in mask generation.
3Measurement precision
If computationally intensive labeling methods are used, then accurate region detection can be achieved, but memory consumption exceeds mobile device capabilities
Solution Approach 1:
The patent segments the image into regions before classification, which reduces the amount of data that needs to be processed and stored in memory. Instead of processing the entire high-resolution image through the classification model, only the segmented regions are fed forward, significantly reducing memory consumption while maintaining detection accuracy.
Solution Approach 2:
The segmentation engine extracts only the relevant regions from the full image before passing them to the classification model. This extraction process removes unnecessary data from memory, reducing overall memory consumption while preserving the accuracy needed for precise region detection and mask generation.
4Manufacturing precision
If full-resolution image processing is performed, then accurate mask borders can be generated, but computational load becomes unsustainable on mobile devices
Solution Approach 1:
The patent segments the image processing task and applies different processing resolutions to different stages. The segmentation engine processes the full-resolution image to identify regions with precise borders, while the classification model processes only the segmented regions at a lower resolution. This approach maintains mask border accuracy while dramatically improving processing efficiency on mobile devices.
Data Source
AI summary
A machine learning system can generate an image mask (e.g., a pixel mask) comprising pixel assignments for pixels. The pixels can be assigned to classes, including, for example, face, clothes, body skin, or hair. The machine learning system can be implemented using a convolutional neural network that is configured to execute efficiently on computing devices having limited resources, such as mobile phones. The pixel mask can be used to more accurately display video effects interacting with a user or subject depicted in the image.