Pedestrian detection method based on deep learning and multi-feature point fusion
A technology of pedestrian detection and deep learning, applied in the direction of instruments, character and pattern recognition, computer components, etc., can solve the problem of mutual occlusion of pedestrians, and achieve the effect of improving accuracy and robustness
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Example Embodiment
[0019] Example
[0020] A pedestrian detection method based on deep learning and multi-feature point fusion, including training phase and detection phase.
[0021] In the training phase, first collect pedestrian images in the application scenario and mark the head and shoulders of the pedestrians in the image, and then use these pedestrian samples for model training. The model training is divided into two steps: 1) Take the pedestrian's head and shoulders image as the training sample and use the Triplet Loss method to train a binary classification model of the pedestrian's head and shoulders; 2) Use the model parameters obtained in step 1) to train Part of the parameters of the pedestrian detection model are initialized by means of "transfer learning". As the model used in the final detection stage, the pedestrian detection model adopts an end-to-end training method, including the functions of candidate area extraction, pedestrian feature extraction and feature classification.
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap