Pedestrian target tracking method based on convolution association network in automatic driving scene
A pedestrian target and associated network technology, applied in image data processing, instrumentation, computing, etc., can solve problems such as high similarity, small proportion of pedestrian targets, and difficulty in pedestrian target detection
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0079] see figure 1 , the present embodiment discloses a pedestrian target tracking method based on a convolutional association network in an automatic driving scene. By using the convolutional association network and the lightweight object detection network to share features, the relevance between objects is captured, thereby To achieve goal tracking, specifically include the following steps:
[0080] A1-1. Obtain a one-stage target detection model, and then after a total of 5 downsamplings, predict on the feature maps of the last three scales. Except for the first downsampling, the ordinary convolution module is used, and the next four downsamplings All the models designed in the multi-scale downsampling module are replaced with separable convolution modules, and the model is used to predict the target frame on the feature map of the last three downsampling, and finally constitute a lightweight pedestrian target detection network;
[0081] A2-1. Predict the target correlati...
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, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com