Anomalous Behaviour Detection using Transfer Learning Algorithm of Series and DAG Network

Published on Oct 1, 2019
· DOI :10.1109/ICSENGT.2019.8906470
Hana' Abd Razak (UiTM: Universiti Teknologi MARA), Ali Abd Almisreb3
Estimated H-index: 3
(International University of Sarajevo)
+ 1 AuthorsNooritawati Md Tahir12
Estimated H-index: 12
(UiTM: Universiti Teknologi MARA)
Transfer learning is highly recommended in image recognition studies due to its ability in leveraging the finest architecture of pre-trained convolution neural networks (CNNs) for instance AlexNet, GoogLeNet and several others in learning new dataset at a faster learning process with smaller input images and could yield better classification rate as well. Hence, this study discussed deep learning with transfer learning approach in recognizing and classifying normal behavior and anomalous behavior referring to housebreaking crime behavior at the gate of the residential unit. Firstly, the dataset of normal behavior and housebreaking crime behavior are acquired. Next, these images are extracted and classified using remodeled AlexNet and GoogLeNet that have been fine-tuned using transfer learning technique. Results attained showed that the classification accuracy for both AlexNet and GoogLeNet are within 97%.
  • References (23)
  • Citations (0)
📖 Papers frequently viewed together
20 Citations
2016IDA: Intelligent Data Analysis
3 Authors (Deepak Soekhoe, ..., Aske Plaat)
10 Citations
78% of Scinapse members use related papers. After signing in, all features are FREE.
#1Kefentse Motshoane (TUT: Tshwane University of Technology)H-Index: 1
#2Chunling Tu (TUT: Tshwane University of Technology)H-Index: 1
Last. Pius A. Owolawi (TUT: Tshwane University of Technology)H-Index: 6
view all 3 authors...
Prohibition signs are commonly used for safety purposes in order to prevent and protect individuals from dangerous situations. These signs are placed in or around areas whereby they are clearly visible to the public. However, the visually impaired cannot visualize such signs. To help them, this paper proposes a system that combines Convolutional Neural Network (CNN) model and Computer Vision (CV) algorithms to detect and recognize prohibition signs in real scenes. The system uses pre-trained Ale...
Recently, several techniques have been explored to detect unusual behaviour in surveillance videos. Nevertheless, few studies leverage features from pre-trained CNNs and none of then present a comparison of features generate by different models. Motivated by this gap, we compare features extracted by four state-of-the-art image classification networks as a way of describing patches from security video frames. We carry out experiments on the Ped1 and Ped2 datasets and analyze the usage of differe...
Nov 1, 2018 in SITIS (Signal-Image Technology and Internet-Based Systems)
#1Geum-Young Son (Kyungsung University)
#2Jang-Sik Park (Kyungsung University)H-Index: 6
view all 4 authors...
This paper proposes a method to detect fire using deep learning. Deep learning networks used in the proposal were AlexNet, GoogLeNet and VGG-16 in three ways. Image input from a closed-circuit television (CCTV) camera is classified into three different states (normal, smoke and flame) and then the network is trained to recognize each corresponding state. The datasets used consist of images coming from three areas which are housing, high-rise building, and mountain where fire incidents occurred. ...
#1H. Wei (UTD: University of Texas at Dallas)
#2M. Laszewski (UTD: University of Texas at Dallas)
Last. Nasser Kehtarnavaz (UTD: University of Texas at Dallas)H-Index: 32
view all 3 authors...
This paper presents a deep learning-based approach to detect and classify persons in video data captured from distances of several miles via a high-power lens video camera. For detection, a set of computationally efficient image processing steps are considered to identify moving areas that contain a person. These areas are then passed onto a convolutional neural network classifier whose convolutional layers consist of the GoogleNet transfer learning. Despite the challenges associated with the vi...
Sep 1, 2018 in ICCCE (International Conference on Computer and Communication Engineering)
#1Subha Fairuz (IIUM: International Islamic University Malaysia)H-Index: 1
#2Mohamed Hadi Habaebi (IIUM: International Islamic University Malaysia)H-Index: 10
Last. Elsheikh Mohamed Ahmed Elsheikh (IIUM: International Islamic University Malaysia)H-Index: 2
view all 3 authors...
Nowadays finger vein-based validation systems are getting extra attraction among other authentication systems due to high security in terms of ensuring data confidentiality. This system works by recognizing patterns from finger vein images and these images are captured using a camera based on near-infrared technology. In this research, we focused finger vein identification system by using our own finger vein dataset, we trained it with transfer learning of AlexNet model and verified by test imag...
1 CitationsSource
May 1, 2018 in IMTC (Instrumentation and Measurement Technology Conference)
#1Feba Thankachan George (CSUF: California State University, Fullerton)H-Index: 1
#2Venkata Sindhoor Preetham Patnam (CSUF: California State University, Fullerton)H-Index: 1
Last. Kiran George (CSUF: California State University, Fullerton)H-Index: 8
view all 3 authors...
As crime rates are increasing all over the world these days, recognizing it in advance will help to reduce it to a significant level. Various studies show that some common signs of non-verbal gestures along with our intellectual judgment can assist to identify a subject with suspicious intentions successfully. A system that utilizes deep learning techniques to detect suspicious non-verbal gestures in real-time is presented in this paper. The system utilizes Jetson TX1 development kit from NVIDIA...
In this paper, the main technologies of foreground detection, feature description and extraction, movement behavior classification and recognition were introduced. Based on optical flow for movement objects detection, optical flow energy image was put forward for movement feature expression and region convolutional neural networks was adopt to choose features and decrease dimension. Then support vector machine classifier was trained and used to classify and recognize actions. After training and ...
#1Suresh DaraH-Index: 1
#2Priyanka TummaH-Index: 1
Deep learning is presently an effective research area in machine learning technique and pattern classification association. This has achieved big success in the areas of application namely computer vision, speech recognition, and NLP. This paper gives the impact of feature extraction that used in a deep learning technique such as Convolutional Neural Network (CNN). The purpose of this paper presents an emerged survey of actual literatures on feature extraction methods since past five years. As t...
2 CitationsSource
#1Azliza Mohd Ali (Lancaster University)H-Index: 2
#2Plamen Angelov (Lancaster University)H-Index: 39
In this paper, we propose a new approach to identify anomalous behaviour based on heterogeneous data and a data fusion technique. There are four types of datasets applied in this study including credit card, loyalty card, GPS, and image data. The first step of the complete framework in this proposed study is to identify the best features for every dataset. Then, the new anomaly detection technique which is recently introduced and known as empirical data analytics (EDA) is applied to detect the a...
2 CitationsSource
#1Mossaad Ben AyedH-Index: 2
#2Mohammed Abid (University of Sfax)H-Index: 27
Video surveillance systems become quickly an essential criteria and requirement for safety systems. The traditional systems based on manual vision should be replaced by intelligent decision. These systems suffer from not only lack of confidence but also from the worst accuracy. Many intelligent systems are proposed in literature to ensure automation and intelligent decision. Suspicious behavior as a particular case of surveillance, presents a great challenge. Problems faced of researchers are th...
Cited By0