Violence Classification Using Support Vector Machine and Deep Transfer Learning Feature Extraction

Karisma,; Imah, Elly Matul; Wintarti, Atik

doi:10.1109/isitia52817.2021.9502253

Cited by 8 publications

(5 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This comparison shows that the accuracies of the proposed models are comparable with the state-of-the-art techniques. (23) 90% 91.66% Soliman et al (24) 95.1% 97% Xingyu et al (25) 95.40% -Karisma et al (26) 92% -Seydi Keceli et al (27) 92.90 98.7 Proposed 96 ± 2% 98 ± 2%…”

Section: Model Evaluationmentioning

confidence: 99%

Violent Video Classification with Transfer Learning Approach Using Inception-V3 and Support Vector Machine

Jaiswal,

Mohod,

Sharma

et al. 2023

IJST

View full text Add to dashboard Cite

Objectives: Research in surveillance systems is growing, with cameras in public places capturing actions for the live surveillance, goal-driven investigation, event forecasting, and intrusion detection. Violent video classification system plays a critical role in development of violence detection system for public security and safety. Such system is useful in identification of violent behaviors, such as fighting or assault. Methods: The Inception-V3 architecture using Convolutional neural networks extracts the informative features from the input video frames. Support Vector Machine is used to select features for classification once the remaining layers of the Inception-V3 model have been frozen. Findings: The datasets used in many contemporary and current innovative techniques, including the Hockey battle dataset and the Movies dataset, are used to train and assess the proposed hybrid model. The experiment findings show that the suggested violence detection algorithm performs well in terms of average metrics, with accuracy, precision, recall, and F-Score being 96 ± 2%, 98 ± 2%, 96 ± 1%, 0.95 respectively. Novelty: Transfer Learning approach is applied which involves lightly retraining pretrained models on different datasets, resulting in improved performance in terms of computational resources and accuracy.

show abstract

Section: Model Evaluationmentioning

confidence: 99%

Violent Video Classification with Transfer Learning Approach Using Inception-V3 and Support Vector Machine

Jaiswal,

Mohod,

Sharma

et al. 2023

IJST

View full text Add to dashboard Cite

show abstract

“…The model utilises different techniques for better performance in detecting violent activities. The paper [22] explored deep TL with SVM for violent video classification. The process involved using VGGNet-16 for feature extraction, followed by SVM for classification using different kernel functions.…”

Section: Related Workmentioning

confidence: 99%

“…The potency of this concept has led to its widespread adoption in various image classification and action recognition tasks. Several studies have employed deep TL models for the automatic detection of violent scenes in videos, such as VGG16 [22], GoogleNet [19], InceptionV3 [17], and MobileNet [8]. However, existing methods employed in video anomaly detection frequently encounter challenges associated with generalisation, which pertains to a model's capability to perform effectively on unseen data.…”

Section: Introductionmentioning

confidence: 99%

“…The methodologies mentioned in [5][6][7][8][9][17][18][19][20]22,25,26,28] faced a shared challenge concerning the integration of new models into the existing framework. These methods require the existing models to be retrained from scratch, resulting in substantial demands on computational resources and time.…”

mentioning

confidence: 99%

“…The existing approaches outlined in references [8, [17][18][19][20]22] employed a single model for feature extraction, thereby overlooking the opportunity to achieve an enhanced feature representation. This limitation arises due to the varying strengths exhibited by different models in capturing specific features or patterns within the data.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Novel Deep Feature Fusion Framework for Multi-Scenario Violence Detection

Jebur,

Hussein,

Hoomod

et al. 2023

Computers

View full text Add to dashboard Cite

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

show abstract

Artificial Intelligent Model for Riot and Violence Detection that Largely Affect Societal Health and Local Healthcare System

Jain,

Bhanodia,

Sethi

2023

Advanced Technologies and Societal Change

View full text Add to dashboard Cite

Violence Classification Using Support Vector Machine and Deep Transfer Learning Feature Extraction

Cited by 8 publications

References 20 publications

Violent Video Classification with Transfer Learning Approach Using Inception-V3 and Support Vector Machine

Violent Video Classification with Transfer Learning Approach Using Inception-V3 and Support Vector Machine

Novel Deep Feature Fusion Framework for Multi-Scenario Violence Detection

Artificial Intelligent Model for Riot and Violence Detection that Largely Affect Societal Health and Local Healthcare System

Contact Info

Product

Resources

About