Lightweight vehicle detection model construction method, system and device

A vehicle detection and lightweight technology, which is applied in the traffic control system of road vehicles, traffic control system, character and pattern recognition, etc., can solve the problems of slow detection speed and low detection accuracy, and achieve the goal of improving detection accuracy and detection speed Effect

Active Publication Date: 2020-12-29
TSINGHUA UNIV
View PDF6 Cites 14 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] At this stage, the vehicle detection model mainly adopts the above-mentioned relat...

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Lightweight vehicle detection model construction method, system and device
  • Lightweight vehicle detection model construction method, system and device
  • Lightweight vehicle detection model construction method, system and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0102] An embodiment of the present invention provides a device for constructing a lightweight vehicle detection model based on deep learning, such as Figure 10 As shown, it includes: a memory 100, a processing 102, and a computer program stored on the memory 100 and operable on the processor 102. When the computer program is executed by the processor 102, the following method steps are implemented:

[0103] Step 101, add a lightweight channel attention module at the end of the bidirectional feature pyramid module of the EfficientNet target detection algorithm EfficientDet model as the backbone network, and learn different channels through the channel attention module with a one-dimensional convolution with a certain step size The relationship between the feature maps; step 101 specifically includes:

[0104] Through the channel attention module, the input C-dimensional feature map is converted into a 1*1*C vector by using global average pooling, and then a one-dimensional co...

Embodiment 2

[0131] An embodiment of the present invention provides a computer-readable storage medium, where a program for realizing information transmission is stored on the computer-readable storage medium, and when the program is executed by the processor 102, the following method steps are implemented:

[0132] Step 101, add a lightweight channel attention module at the end of the bidirectional feature pyramid module of the EfficientNet target detection algorithm EfficientDet model as the backbone network, and learn different channels through the channel attention module with a one-dimensional convolution with a certain step size The relationship between the feature maps; step 101 specifically includes:

[0133] Through the channel attention module, the input C-dimensional feature map is converted into a 1*1*C vector by using global average pooling, and then one-dimensional convolution with a convolution kernel size of k is performed, and then learned through the Sigmoid function outpu...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a lightweight vehicle detection model construction method, system and device. The method comprises the following steps of: adding a lightweight channel attention module at thetail end of a bidirectional feature pyramid module of a target detection algorithm EfficientDet model with an EfficientNet backbone network, and learning a relationship of feature maps among differentchannels through the channel attention module by one-dimensional convolution with a certain step length; carrying out positive and negative sample selection by adopting an adaptive training sample selection algorithm, and training the EfficientDet model; training a high-precision model by using a large model in an original data set by adopting a distillation learning method; and then compressingthe trained large model to construct a smaller model, and guiding the small model to carry out distillation learning of an intermediate feature layer, distillation learning of a target classificationresult and result distillation learning of target frame regression prediction by using the large model as a teacher network to obtain a final lightweight vehicle detection model.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, in particular to a method, system and device for constructing a lightweight vehicle detection model. Background technique [0002] Before the emergence of deep learning technology, computer vision tasks mainly used traditional manual features combined with classifier methods for object recognition and detection. These traditional machine learning methods had problems such as slow speed and low precision. As the research of deep learning technology becomes more and more mature, coupled with the massive multimedia data accumulated in the Internet and the computing power of various computing devices continue to improve, it is possible to apply deep learning technology to computer vision tasks and achieve good results. Currently, the mainstream methods of computer vision tasks such as object detection, object classification, and object tracking are all based on deep learning. ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G08G1/017G06K9/00G06K9/62G06K9/46
CPCG08G1/017G06V20/54G06V10/40G06V2201/08G06F18/217G06F18/214
Inventor 丁贵广冷宸宇
Owner TSINGHUA UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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