Face and Hand Association Detection via Multi-Scale Feature Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to directly associate human faces and hands in images due to their spatial separation, requiring key point technology and object detection which can be complex and less accurate.
Innovation Solution
A human face and hand association detecting method that involves acquiring an image of a human object, performing feature extraction to obtain multiple scale feature maps, and then fusing these feature maps to enhance feature accuracy, allowing for the detection of associated human face and hand positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If key point technology and object detection technology are used to associate human face and hand, then association can be achieved, but the device complexity increases and measurement precision decreases
Solution Approach 1:
The detection process is segmented into distinct stages: face detection, hand detection, and association detection. Each stage uses specialized detectors optimized for their specific task, allowing the system to achieve high accuracy without requiring a single complex all-purpose detector. The association module then connects detected faces and hands using spatial relationship analysis.
Solution Approach 2:
An association detection module is introduced as an intermediary component that bridges face detection and hand detection results. This module analyzes spatial relationships and uses machine learning models to determine whether detected faces and hands belong to the same person, improving association accuracy without increasing the complexity of the base detection systems.
2Measurement precision
If multi-scale feature fusion is performed, then measurement precision of feature information improves, but use of energy and computation increases
Solution Approach 1:
The system applies different processing intensities to different regions and scales of the input image. Multi-scale feature fusion is performed selectively at critical detection stages and regions, rather than uniformly across the entire image. This allows the system to maintain high feature information accuracy while reducing overall computational energy consumption by focusing processing resources where they are most needed.
Data Source
AI summary
The present disclosure relates to a method and apparatus for association detection for a human face and a human hand, an electronic apparatus and a storage medium. The method comprises: acquiring a first image, the first image being an image of a human object; performing feature extraction on the first image to obtain a first feature map of multiple dimensions; performing a feature fusion processing on the first feature map of multiple dimensions to obtain a second feature map of multiple dimensions, the dimensions of the second feature map corresponding to the dimensions of the first feature map in a one-to-one manner; detecting, on the basis of the obtained second feature map of multiple dimensions, a human face position and a human hand position associated with the same human object in the first image. The embodiment of the present disclosure enables easy and convenient association detection for human face and human hand.