Neural Network Operator Segmentation for Image Detection Speed
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing neural network models for image detection require large amounts of memory to process huge input and output data, leading to slow image detection performance.
Innovation Solution
The method involves mapping the neural network model to a singly-linked list, calculating output and input numbers for each node, establishing a vector of data pairs, and determining operator subsets based on these data pairs to reduce the amount of data recorded and processed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If neural network models process huge input and output data between operators, then image detection accuracy is maintained, but memory usage increases and detection speed decreases
Solution Approach 1:
The patent segments the neural network operators into distinct groups and tracks only the necessary input-output data relationships between these segments. By organizing operators into a structured framework where only relevant data flows are recorded, the system reduces overall data volume while maintaining detection accuracy and improving processing speed.
Solution Approach 2:
The patent extracts and records only the essential input-output data relationships between operators rather than recording all data. By identifying and extracting only the necessary data flow information, the system significantly reduces memory usage and processing requirements while preserving the accuracy needed for effective image detection.
2Reliability
If neural network models include many operators with huge data amounts, then detection accuracy is maintained, but memory consumption increases
Solution Approach 1:
The patent extracts and records only the essential input-output data relationships between operators rather than recording all data. By identifying and extracting only the necessary data flow information, the system significantly reduces memory usage while preserving the accuracy needed for effective image detection.
Solution Approach 2:
The patent creates a universal data recording framework that handles multiple operators and data types through a single structured approach. By using a unified method to track input-output relationships across different operator types, the system reduces redundancy and memory consumption while maintaining comprehensive detection capability.
3Productivity
If all operator data is recorded for neural network processing, then detection completeness is maintained, but processing time increases
Solution Approach 1:
The patent extracts and records only the essential input-output data relationships between operators rather than recording all data. By identifying and extracting only the necessary data flow information, the system significantly reduces processing time while maintaining detection accuracy.
Solution Approach 2:
The patent performs preliminary analysis to identify which data relationships are necessary before actual processing begins. By pre-determining the essential input-output connections and organizing operators accordingly, the system avoids unnecessary data processing steps and reduces overall processing time while maintaining completeness.
Data Source
AI summary
An image detection and classification method for faster detection and classification by a neural network is applied to a neural network model including n operators. The neural network model is mapped to a singly-linked list. A vector of data pairs corresponding to the singly-linked list is established, and operator subsets of the neural network model are determined according to the vector. A target image is input to the neural network model, and input data and output data of the operator subset are recorded. The neural network model is applied to detect the target image, and a detection result is output according to the input data and the output data. The method can detect images quickly.