Leveraging YOLOv5s with optimization‐based effective anomaly detection in pedestrian walkways DOI
Allabaksh Shaik, Shaik Mahaboob Basha

Expert Systems, Journal Year: 2024, Volume and Issue: unknown

Published: May 23, 2024

Abstract Currently, video surveillance is generally used to safeguard safety in public places like railway stations, traffic signals, malls, and so on. Video anomaly recognition localization are the main components of intelligent method. refers procedure spatiotemporal abnormal design existing video. A task classification anomalies that occur it thefts, crimes, forth. Also, pedestrian walkways has enlarged major attention among computer vision (CV) groups improve protection. The current developments Deep Learning (DL) methods have great dissimilar procedures image classification, object recognition, This study designs an Optimal for Effective Anomaly Detection Pedestrian Walkways (ODL‐EADPW) model. ODL‐EADPW technique employs a fine‐tuned DL model identification pedestrians walkways. In technique, pre‐processing primarily involved two stages median filtering (MF) based noise removal adaptive histogram equalization (AHE)‐based contrast enhancement. For detection walkways, uses YOLOv5s with EfficientRep as backbone network. To enhance results stochastic gradient descent (SGD) optimizer was employed perfect hyperparameters performance evaluation methodology implemented on UCSD dataset. An extensive comparison stated gains effectual over other models terms different measures.

Language: Английский

Leveraging YOLOv5s with optimization‐based effective anomaly detection in pedestrian walkways DOI
Allabaksh Shaik, Shaik Mahaboob Basha

Expert Systems, Journal Year: 2024, Volume and Issue: unknown

Published: May 23, 2024

Abstract Currently, video surveillance is generally used to safeguard safety in public places like railway stations, traffic signals, malls, and so on. Video anomaly recognition localization are the main components of intelligent method. refers procedure spatiotemporal abnormal design existing video. A task classification anomalies that occur it thefts, crimes, forth. Also, pedestrian walkways has enlarged major attention among computer vision (CV) groups improve protection. The current developments Deep Learning (DL) methods have great dissimilar procedures image classification, object recognition, This study designs an Optimal for Effective Anomaly Detection Pedestrian Walkways (ODL‐EADPW) model. ODL‐EADPW technique employs a fine‐tuned DL model identification pedestrians walkways. In technique, pre‐processing primarily involved two stages median filtering (MF) based noise removal adaptive histogram equalization (AHE)‐based contrast enhancement. For detection walkways, uses YOLOv5s with EfficientRep as backbone network. To enhance results stochastic gradient descent (SGD) optimizer was employed perfect hyperparameters performance evaluation methodology implemented on UCSD dataset. An extensive comparison stated gains effectual over other models terms different measures.

Language: Английский

Citations

0