Novel Deep Feature Fusion Framework for Multi-Scenario Violence Detection

Computers, Journal Year: 2023, Volume and Issue: 12(9), P. 175 - 175

Published: Sept. 5, 2023

Detecting violence in various scenarios is a difficult task that requires high degree of generalisation. This includes fights different environments such as schools, streets, and football stadiums. However, most current research on detection focuses single scenario, limiting its ability to generalise across multiple scenarios. To tackle this issue, paper offers new multi-scenario framework operates two environments: fighting locations rugby has three main steps. Firstly, it uses transfer learning by employing pre-trained models from the ImageNet dataset: Xception, Inception, InceptionResNet. approach enhances generalisation prevents overfitting, these have already learned valuable features large diverse dataset. Secondly, combines extracted through feature fusion, which improves representation performance. Lastly, concatenation step first scenario with second train machine classifier, enabling classifier both highly flexible, can incorporate without requiring training scratch additional The Fusion model, incorporates fusion models, obtained an accuracy 97.66% RLVS dataset 92.89% Hockey Concatenation model accomplished 97.64% 92.41% datasets just classifier. allows for classification violent within Furthermore, not limited be adapted tasks.

Language: Английский

---

Enhancing Video Anomaly Detection Using a Transformer Spatiotemporal Attention Unsupervised Framework for Large Datasets

Algorithms, Journal Year: 2024, Volume and Issue: 17(7), P. 286 - 286

Published: July 1, 2024

This work introduces an unsupervised framework for video anomaly detection, leveraging a hybrid deep learning model that combines vision transformer (ViT) with convolutional spatiotemporal relationship (STR) attention block. The proposed addresses the challenges of detection in surveillance by capturing both local and global relationships within frames, task traditional neural networks (CNNs) often struggle due to their localized field view. We have utilized pre-trained ViT as encoder feature extraction, which is then processed STR block enhance among objects videos. novelty this utilizing detect anomalies effectively large heterogeneous datasets, important thing given diverse environments scenarios encountered real-world surveillance. was evaluated on three benchmark i.e., UCSD-Ped2, CHUCK Avenue, ShanghaiTech. demonstrates model’s superior performance detecting compared state-of-the-art methods, showcasing its potential significantly automated systems achieving area under receiver operating characteristic curve (AUC ROC) values 95.6, 86.8, 82.1. To show effectiveness extra-large we trained subset huge contemporary CHAD dataset contains over 1 million AUC ROC 71.8 64.2 CHAD-Cam 2, respectively, outperforms techniques.

Language: Английский

---

VEAD: Variance profile Exploitation for Anomaly Detection in real-time IoT data streaming

Internet of Things, Journal Year: 2023, Volume and Issue: 25, P. 100994 - 100994

Published: Nov. 17, 2023

Language: Английский

---

Towards Transfer Learning Based Human Anomaly Detection in Videos

Lecture notes in networks and systems, Journal Year: 2024, Volume and Issue: unknown, P. 411 - 425

Published: Jan. 1, 2024

Language: Английский

---

A Region based Salient Stacking Optimized Detector (ReSOD) For an Effective Anomaly Detection and Video Tracking in Surveillance Systems

Neurocomputing, Journal Year: 2024, Volume and Issue: unknown, P. 129281 - 129281

Published: Dec. 1, 2024

Language: Английский

---