Pedestrian Re-Identification Network Optimization via Feature Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current pedestrian re-identification methods face challenges in extracting discriminative features due to pedestrian pose diversity and background diversity in image datasets, leading to suboptimal identity recognition performance.
Innovation Solution
A network optimization method that involves obtaining image sample groups with paired images of the same and different objects, extracting identity and attribute features, performing feature exchange processing, and adjusting neural network parameters based on loss values to enhance recognition precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If decomposition features are extracted by using a deep neural network for identity recognition, then identity recognition capability is improved, but the network becomes sensitive to pose diversity and background diversity
Solution Approach 1:
The patent segments the feature extraction process into two independent parts: identity features and attribute features. The identity feature extraction network focuses solely on extracting identity-related information, while the attribute feature extraction network handles pose, background, and other non-identity attributes. This segmentation allows each network to specialize, preventing the identity recognition network from being confounded by pose and background variations.
Solution Approach 2:
The patent extracts and removes attribute features (pose, background, etc.) from the feature representation space. By separating attribute features from identity features through independent extraction networks, the identity recognition system can focus purely on identity information without the noise and interference from pose and background diversity.
2Measurement precision
If feature exchange processing is performed to separate identity and attribute features, then recognition precision is improved, but computational complexity increases
Solution Approach 1:
The computational complexity is managed through segmentation of the feature extraction process. Instead of processing all features together in a single complex network, the system divides the task into two separate extraction networks (identity and attribute), each handling a specific subset of features. This modular approach reduces the computational burden on any single network while maintaining the benefits of comprehensive feature separation.
Data Source
AI summary
The present disclosure relates to a network optimization method and apparatus, an image processing method and apparatus, and a storage medium. The network optimization method includes: obtaining an image sample group; obtaining a first feature and a second feature of an image in the image sample group, and obtaining a first classification result by using the first feature of the image; performing feature exchange processing on an image pair in the image sample group to obtain a new image pair; obtaining a first loss value of the first classification result, a second loss value of the new image pair, and a third loss value of first features and second features of the new image pair in a preset manner; and adjusting parameters of a neural network at least according to the first loss value, the second loss value, and the third loss value until a preset requirement is met.