An intelligent system to detect human suspicious activity using deep neural networks

Ramachandran, Sumalatha; Palivela, Lakshmi Harika

doi:10.3233/jifs-179003

Cited by 13 publications

(7 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These methods involve descriptors for low-level representations of video features and use 2D cells or 3D cubes of each frame interest point. After feature description, classifiers are used for classification [25,26].…”

Section: Non-object Centric Based Methods For Violence Detectionmentioning

confidence: 99%

See 1 more Smart Citation

Real-Time Violent Action Recognition Using Key Frames Extraction and Deep Learning

Ahmed¹,

Ramzan²,

Khan³

et al. 2021

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

Violence recognition is crucial because of its applications in activities related to security and law enforcement. Existing semi-automated systems have issues such as tedious manual surveillances, which causes human errors and makes these systems less effective. Several approaches have been proposed using trajectory-based, non-object-centric, and deep-learning-based methods. Previous studies have shown that deep learning techniques attain higher accuracy and lower error rates than those of other methods. However, the their performance must be improved. This study explores the state-of-the-art deep learning architecture of convolutional neural networks (CNNs) and inception V4 to detect and recognize violence using video data. In the proposed framework, the keyframe extraction technique eliminates duplicate consecutive frames. This keyframing phase reduces the training data size and hence decreases the computational cost by avoiding duplicate frames. For feature selection and classification tasks, the applied sequential CNN uses one kernel size, whereas the inception v4 CNN uses multiple kernels for different layers of the architecture. For empirical analysis, four widely used standard datasets are used with diverse activities. The results confirm that the proposed approach attains 98% accuracy, reduces the computational cost, and outperforms the existing techniques of violence detection and recognition.

show abstract

Section: Non-object Centric Based Methods For Violence Detectionmentioning

confidence: 99%

“…2 lists the trajectory-based approaches used to detect violence. OMEGA equations SVM 90% [30] Gaussian and fuzzy K mean K-NN 95% [25] Motion blob SVM 92%…”

Section: Non-object Centric Based Methods For Violence Detectionmentioning

confidence: 99%

Real-Time Violent Action Recognition Using Key Frames Extraction and Deep Learning

Ahmed¹,

Ramzan²,

Khan³

et al. 2021

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

show abstract

“…The deep learning model proposed in most of the 24 applied research articles in the sample was the convolutional neural networks (CNN) model, followed by the Gaussian mixture model (GMM). Nine sampled applied research manuscripts introduced a CNNbased model [15,29,31,33,[35][36][37][38][39], whereas seven applied research articles presented a GMMbased model [40][41][42][43]. These models are briefly detailed as follows.…”

Section: What Deep Learning Models Have Been Used To Detect Anti-soci...mentioning

confidence: 99%

“…Ramachandran and Palivela [36] proposed a framework aimed at detecting suspicious human behaviour in surveillance videos and distinguishing it from normal activities. The model utilised a CNN to extract features from optical flow slices and pre-trained the activities based on real-time data.…”

Section: Cnn-based Deep Learning Models Sampledmentioning

confidence: 99%

The Role of Deep Learning Models in the Detection of Anti-Social Behaviours towards Women in Public Transport from Surveillance Videos: A Scoping Review

Papini,

Iqbal,

Barthelemy

et al. 2023

Safety

View full text Add to dashboard Cite

Increasing women’s active participation in economic, educational, and social spheres requires ensuring safe public transport environments. This study investigates the potential of machine learning-based models in addressing behaviours impacting the safety perception of women commuters. Specifically, we conduct a comprehensive review of the existing literature concerning the utilisation of deep learning models for identifying anti-social behaviours in public spaces. Employing a scoping review methodology, our study synthesises the current landscape, highlighting both the advantages and challenges associated with the automated detection of such behaviours. Additionally, we assess available video and audio datasets suitable for training detection algorithms in this context. The findings not only shed light on the feasibility of leveraging deep learning for recognising anti-social behaviours but also provide critical insights for researchers, developers, and transport operators. Our work aims to facilitate future studies focused on the development and implementation of deep learning models, enhancing safety for all passengers in public transportation systems.

show abstract

“…In [48], an intelligent automatic system to detect behavior of the human in public places is presented. The framework to detect suspicious human behavior as well as tracking of human who is doing some unusual activity such as fighting and threatening actions and also distinguishing the human normal activities from the suspicious behavior is proposed.…”

mentioning

confidence: 99%

Intelligent, smart and scalable cyber-physical systems

Vijayakumar

Subramaniyaswamy

Abawajy

et al. 2019

IFS

View full text Add to dashboard Cite

The integration of human physical processes with the computation has created a new paradigm called Cyber-Physical Systems (CPS). The CPSs control the physical processes by utilizing the computational intelligence to acquire the deep knowledge of the monitored environment. Hence the CPSs are designed to be intelligent to provide highly accurate decisions and appropriate actions promptly. The rapidly growing interconnections between the virtual and physical worlds and the development of new intelligent techniques have created new opportunities for the research for next-generation CPS, that is intelligent cyber-physical systems (iCPS). The iCPS are large-scale software intensive and pervasive systems, which by combining various data sources (both from physical objects and virtual components) and applying intelligence techniques, can efficiently manage real-world processes and offers a broad range of novel applications and services. By equipping physical objects with interfaces to the virtual world, and incorporating intelligent mechanisms to leverage collaboration between these objects, the boundaries between the physical and virtual worlds become blurred. Interactions occurring in the physical world are capable of changing the processinga behavior in the virtual world, in a causal relationship

show abstract

An intelligent system to detect human suspicious activity using deep neural networks

Cited by 13 publications

References 8 publications

Real-Time Violent Action Recognition Using Key Frames Extraction and Deep Learning

Real-Time Violent Action Recognition Using Key Frames Extraction and Deep Learning

The Role of Deep Learning Models in the Detection of Anti-Social Behaviours towards Women in Public Transport from Surveillance Videos: A Scoping Review

Intelligent, smart and scalable cyber-physical systems

Contact Info

Product

Resources

About