scholarly journals Anomaly Detection in Videos Using Two-Stream Autoencoder with Post Hoc Interpretability

2021 ◽  
Vol 2021 ◽  
pp. 1-15
Author(s):  
Jiangfan Feng ◽  
Yukun Liang ◽  
Lin Li
Keyword(s):  
Anomaly Detection ◽  
Video Sequence ◽  
Feature Learning ◽  
Learning Approaches ◽  
Surveillance Video ◽  
Efficient Detection ◽  
Post Hoc ◽  
Decision Boundaries ◽  
Object Level ◽  
Abnormal Events

The growing interest in deep learning approaches to video surveillance raises concerns about the accuracy and efficiency of neural networks. However, fast and reliable detection of abnormal events is still a challenging work. Here, we introduce a two-stream approach that offers an autoencoder-based structure for fast and efficient detection to facilitate anomaly detection from surveillance video without labeled abnormal events. Furthermore, we present post hoc interpretability of feature map visualization to show the process of feature learning, revealing uncertain and ambiguous decision boundaries in the video sequence. Experimental results on Avenue, UCSD Ped2, and Subway datasets show that our method can detect abnormal events well and explain the internal logic of the model at the object level.

scholarly journals Anomaly Detection in Human Behavior using Video Surveillance

2019 ◽  
Vol 9 (2) ◽  
pp. 328-332
Keyword(s):  
Anomaly Detection ◽  
Video Sequence ◽  
Good Accuracy ◽  
Review Paper ◽  
Test Sample ◽  
Reconstruction Error ◽  
Surveillance Video ◽  
Huge Number ◽  
Video Cameras ◽  
Do So

Conventional static surveillance has proved to be quite ineffective as the huge number of cameras to keep an eye on most often outstrips the monitor’s ability to do so. Furthermore, the amount of focus needed to constantly monitor the surveillance video cameras is often overbearing. The review paper focuses on solving the problem of anomaly detection in video sequence through semi-supervised techniques. Each video is defined as sequence of frames. The model is trained with goal to minimize the reconstruction error which later on is used to detect anomaly in the test sample videos. The model was trained and tested on most commonly used benchmarking datasetAvenue dataset. Experiment results confirm that the model detects anomaly in a video with a reasonably good accuracy in presence of some noise in dataset

scholarly journals Anomaly detection in discrete manufacturing using self-learning approaches

Procedia CIRP ◽  
2019 ◽  
Vol 79 ◽  
pp. 313-318 ◽  
Author(s):  
Benjamin Lindemann ◽  
Fabian Fesenmayr ◽  
Nasser Jazdi ◽  
Michael Weyrich
Keyword(s):  
Anomaly Detection ◽  
Learning Approaches ◽  
Discrete Manufacturing ◽  
Self Learning

scholarly journals Survey Analysis of Robust and Real-Time Multi-Lane and Single Lane Detection in Indian Highway Scenarios

2021 ◽  
Vol 309 ◽  
pp. 01117
Author(s):  
A. Sai Hanuman ◽  
G. Prasanna Kumar
Keyword(s):  
Neural Network ◽  
System Integration ◽  
Large Scale ◽  
Feature Learning ◽  
Semantic Segmentation ◽  
Lane Detection ◽  
Learning Approaches ◽  
Survey Analysis ◽  
Lane Recognition ◽  
Continuous Frames

Studies on lane detection Lane identification methods, integration, and evaluation strategies square measure all examined. The system integration approaches for building a lot of strong detection systems are then evaluated and analyzed, taking into account the inherent limits of camera-based lane detecting systems. Present deep learning approaches to lane detection are inherently CNN's semantic segmentation network the results of the segmentation of the roadways and the segmentation of the lane markers are fused using a fusion method. By manipulating a huge number of frames from a continuous driving environment, we examine lane detection, and we propose a hybrid deep architecture that combines the convolution neural network (CNN) and the continuous neural network (CNN) (RNN). Because of the extensive information background and the high cost of camera equipment, a substantial number of existing results concentrate on vision-based lane recognition systems. Extensive tests on two large-scale datasets show that the planned technique outperforms rivals' lane detection strategies, particularly in challenging settings. A CNN block in particular isolates information from each frame before sending the CNN choices of several continuous frames with time-series qualities to the RNN block for feature learning and lane prediction.

A Comparative Review of Various Machine Learning Approaches for Improving the Performance of Stego Anomaly Detection

2018 ◽  
pp. 351-371 ◽  
Author(s):  
Hemalatha Jeyaprakash ◽  
KavithaDevi M. K. ◽  
Geetha S.
Keyword(s):  
Machine Learning ◽  
Anomaly Detection ◽  
Detection Rate ◽  
High Dimensional ◽  
Learning Approaches ◽  
Learning Tools ◽  
High Detection Rate ◽  
Comparative Review ◽  
High Detection

In recent years, steganalyzers are intelligently detecting the stego images with high detection rate using high dimensional cover representation. And so the steganographers are working towards this issue to protect the cover element dependency and to protect the detection of hiding secret messages. Any steganalysis algorithm may achieve its success in two ways: 1) extracting the most sensitive features to expose the footprints of message hiding; 2) designing or building an effective classifier engine to favorably detect the stego images through learning all the stego sensitive features. In this chapter, the authors improve the stego anomaly detection using the second approach. This chapter presents a comparative review of application of the machine learning tools for steganalysis problem and recommends the best classifier that produces a superior detection rate.

Detection of Classical Cipher Types with Feature-Learning Approaches

2021 ◽  
pp. 152-164
Author(s):  
Ernst Leierzopf ◽  
Vasily Mikhalev ◽  
Nils Kopal ◽  
Bernhard Esslinger ◽  
Harald Lampesberger ◽  
...  
Keyword(s):  
Feature Learning ◽  
Learning Approaches

scholarly journals Hybridizing Feature Selection and Feature Learning Approaches in QSAR Modeling for Drug Discovery

2017 ◽  
Vol 7 (1) ◽  
Author(s):  
Ignacio Ponzoni ◽  
Víctor Sebastián-Pérez ◽  
Carlos Requena-Triguero ◽  
Carlos Roca ◽  
María J. Martínez ◽  
...  
Keyword(s):  
Feature Selection ◽  
Drug Discovery ◽  
Feature Learning ◽  
Learning Approaches ◽  
Qsar Modeling

scholarly journals Machine Learning approaches for Anomaly Detection in Multiphase Flow Meters

2019 ◽  
Vol 52 (11) ◽  
pp. 212-217 ◽  
Author(s):  
Tommaso Barbariol ◽  
Enrico Feltresi ◽  
Gian Antonio Susto
Keyword(s):  
Machine Learning ◽  
Anomaly Detection ◽  
Multiphase Flow ◽  
Learning Approaches ◽  
Flow Meters

scholarly journals Road Anomaly Detection Through Deep Learning Approaches

IEEE Access ◽  
2020 ◽  
Vol 8 ◽  
pp. 117390-117404
Author(s):  
Dawei Luo ◽  
Jianbo Lu ◽  
Gang Guo
Keyword(s):  
Deep Learning ◽  
Anomaly Detection ◽  
Learning Approaches
Sign in / Sign up

Export Citation Format

Share Document