Multi-Task Image Processing with Spatial-Channel Feature Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies in computer vision, particularly for autonomous driving, face challenges in achieving high accuracy and real-time performance for multiple interrelated tasks.
Innovation Solution
An electronic system utilizing an AI model that extracts image features, generates spatial and channel features, fuses them, and performs specialized image processing tasks using customized task features, enhanced by spatial and channel attention mechanisms, to enhance processing speed and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple image processing tasks are performed using separate processing pipelines, then each task can achieve high accuracy, but the computational load increases and real-time performance deteriorates
Solution Approach 1:
The patent merges multiple image processing tasks into a unified neural network architecture that shares common feature extraction layers. The system processes multiple tasks (e.g., object detection, segmentation, tracking) simultaneously through a single integrated model, reducing redundant computations while maintaining accuracy through task-specific output heads
Solution Approach 2:
The patent implements a universal base model that serves multiple image processing functions. The shared feature extraction backbone provides common representations that can be adapted to different tasks through configurable task modules, enabling one system to perform multiple specialized functions efficiently
2Measurement precision
If multiple image processing tasks are performed using separate processing pipelines, then each task can achieve high accuracy, but the computational complexity increases
Solution Approach 1:
The patent combines multiple task-specific processing streams into a single unified architecture. By merging the feature extraction backbones and sharing computational resources across tasks, the system reduces overall computational complexity while maintaining the ability to perform multiple specialized functions through modular task heads
3Measurement precision
If feature extraction is performed separately for each image processing task, then each task can achieve high precision, but the computational load increases and processing time increases
Solution Approach 1:
The patent performs feature extraction once at the beginning of the processing pipeline, creating a shared representation that is then reused by multiple task-specific modules. This preliminary feature extraction eliminates redundant computations that would occur if each task extracted features independently, significantly reducing processing time while preserving precision through task-adaptive processing of the shared features
Data Source
AI summary
A method, device, and system with image processing task performance are provided. The electronic system includes one or more processors, and a memory storing code for performing at least two image processing tasks with respect to an image, wherein execution of the code by the one or more processors causes the one or more processors to extract an image feature of the image, respectively generate a spatial feature and a channel feature dependent on the image feature, generate a fused feature, for the image, dependent on the spatial feature and the channel feature, and generate respective results of the at least two image processing tasks based on a corresponding customized task feature for each of the at least two image processing tasks generated using the fused feature.