Research on Lightweight Infrared Pedestrian Detection Model Algorithm for Embedded Platform
Authors:
Open Access
- 30 November 2021
- journal article
- retracted article
- Published by Hindawi Limited in Security and Communication Networks
- Vol. 2021, 1-7
Abstract
Due to the limitation of energy consumption and power consumption, the embedded platform cannot meet the real-time requirements of the far-infrared image pedestrian detection algorithm. To solve this problem, this paper proposes a new real-time infrared pedestrian detection algorithm (RepVGG-YOLOv4, Rep-YOLO), which uses RepVGG to reconstruct the YOLOv4 backbone network, reduces the amount of model parameters and calculations, and improves the speed of target detection; using space spatial pyramid pooling (SPP) obtains different receptive field information to improve the accuracy of model detection; using the channel pruning compression method reduces redundant parameters, model size, and computational complexity. The experimental results show that compared with the YOLOv4 target detection algorithm, the Rep-YOLO algorithm reduces the model volume by 90, the floating-point calculation is reduced by 93.4, the reasoning speed is increased by 4 times, and the model detection accuracy after compression reaches 93.25.Keywords
Funding Information
- JiangSu Collaborative Innovation Center for Building Energy Saving and Construction Technology (SJXTY1603, 186130061, KC21108, KC21335, 2020-R-84360)
This publication has 12 references indexed in Scilit:
- Rethinking RGB-D Salient Object Detection: Models, Data Sets, and Large-Scale BenchmarksIEEE Transactions on Neural Networks and Learning Systems, 2020
- Asymmetric Non-Local Neural Networks for Semantic Segmentation2019
- Dim and small target detection based on feature mapping neural networksJournal of Visual Communication and Image Representation, 2019
- Semi-Supervised Sparse Representation Based Classification for Face Recognition With Insufficient Labeled SamplesIEEE Transactions on Image Processing, 2017
- Deep Residual Learning for Image Recognition2016
- You Only Look Once: Unified, Real-Time Object Detection2016
- DeepFlow: Large Displacement Optical Flow with Deep Matching2013
- Infrared Patch-Image Model for Small Target Detection in a Single ImageIEEE Transactions on Image Processing, 2013
- Saliency filters: Contrast based filtering for salient region detection2012
- Facial expression recognition from near-infrared videosImage and Vision Computing, 2011