Gesture Recognition Using Reference Vectors for Faster Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing gesture recognition technologies face high computational complexity and inefficiency due to the need to process and analyze a large number of calculation operations, especially in deep learning models, which are not effective in solving the problem of computational complexity and inefficiency in the field of gesture recognition technology.
Innovation Solution
A method and device for calculating similarities between a gesture feature vector and reference vectors are calculated to determine the gesture category.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deep learning models are used for gesture recognition, then gesture recognition accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent segments the gesture recognition process into two distinct phases: an offline training phase where the deep learning model is built and stored, and an online recognition phase where only simple similarity comparison is performed. This segmentation allows the complex model construction to be done once offline, while the online phase uses simplified operations to maintain low computational complexity during actual gesture recognition.
Solution Approach 2:
The patent performs preliminary action by pre-training the deep learning model offline to generate and store reference feature vectors for various gesture categories. This preliminary training phase prepares all necessary computational resources and reference data in advance, so that during online recognition, only simple vector similarity comparisons are needed, significantly reducing real-time computational complexity.
2Measurement precision
If deep learning models process large amounts of gesture image data, then gesture recognition accuracy is improved, but gesture recognition speed decreases
Solution Approach 1:
The patent divides the gesture recognition system into offline model training and online similarity comparison. The complex deep learning processing is segmented into the offline phase where reference vectors are pre-computed, while the online phase uses segmented, simplified vector comparison operations that are computationally lightweight and fast.
Solution Approach 2:
The patent performs preliminary action by pre-processing gesture image data offline to train the deep learning model and generate reference feature vectors. This preliminary processing extracts and stores essential gesture characteristics in advance, so that online recognition only requires comparing new gestures against these pre-extracted features using fast similarity metrics, dramatically improving recognition speed.
3Measurement precision
If more calculation operations are performed on gesture feature data, then gesture recognition accuracy is improved, but gesture recognition efficiency decreases
Solution Approach 1:
The patent segments computational operations into offline feature extraction and model training, followed by online similarity comparison. This segmentation moves computationally intensive operations to the offline phase, while the online phase uses efficient, minimal calculation operations to maintain high gesture recognition efficiency without sacrificing accuracy.
Solution Approach 2:
The patent performs preliminary action by pre-extracting gesture features and pre-training the classification model offline. This preliminary computation establishes reference feature vectors and model parameters in advance, so that online recognition only requires simple similarity calculations, maximizing gesture recognition efficiency while maintaining high accuracy through the pre-computed reference data.
Data Source
AI summary
A gesture recognition method, a gesture recognition device, an electronic device and a computer-readable storage medium are provided. The gesture recognition method includes steps: obtaining a reference feature vector set including M gesture categories and a gesture category set including M reference feature vectors, where each of the reference feature vectors is obtained by performing vector fusion on initial feature vectors of N sample images of each of the gesture categories, and the initial feature vectors are obtained by performing hand feature extraction on the sample images; performing hand feature extraction on the image to be recognized to obtain a gesture feature vector; and determining a target gesture category of the image to be recognized based on similarities between the gesture feature vector and the M reference feature vectors. The gesture recognition method reduces computational complexity and improves gesture recognition efficiency