Detection of unusual behavior is an important topic in signal and image processing. Because of the topic's complexity, addressing it as a solely RGB video analysis problem raises significant challenges. This has resulted in approaches that aim at exploiting different data modalities that can overcome the inherent restrictions of unimodal techniques. Moreover, the classification outcome of such approaches is affected not only by the input data, but also by previous classification history. To this end, this paper introduces a novel deep-NARMA filter that extends a typical CNN architecture, and endows it with autoregressive moving average behavior. In addition, it incorporates a data fusion framework that supplements RGB video streams, with thermal capturing and information about the distortion of WiFi signal reflectance. Experimental results indicate a better performance compared to conventional as well as deep learning approaches.
The advancement of sensing technologies coupled with the rapid progress in big data analysis has ushered in a new era in intelligent transport and smart city applications. In this context, transportation mode detection (TMD) of mobile users is a field that has gained significant traction in recent years. In this paper, we present a deep learning approach for transportation mode detection using multimodal sensor data elicited from user smartphones. The approach is based on long short-term Memory networks and Bayesian optimization of their parameters. We conducted an extensive experimental evaluation of the proposed approach, which attains very high recognition rates, against a multitude of machine learning approaches, including state-of-the-art methods. We also discuss issues regarding feature correlation and the impact of dimensionality reduction.
In this paper we present a video summarization method that extracts key-frames from industrial surveillance videos, thus dramatically reducing the number of frames without significant loss of semantic content. We propose to use the produced summaries as training set for neural network based Evaluative Rectification. Evaluative Rectification is a method that exploits an expert user's feedback regarding the correctness of an activity recognition framework on part of the data in order to enhance future classification results. The size of the training sample set usually depends on the topology of the network and on the complexity of the environment and activities observed. However, as is shown by the experiments conducted in a real-world industrial activity recognition dataset, using a much smaller but representative sample stemming from our summarization technique leads to significantly higher accuracy rates than those attained by a same size but randomly chosen set. To obtain comparable improvement in accuracy without the summarization technique, the experiments show that a far larger training sample set is needed, therefore requiring significantly increased human resources and computational cost.
Unlike any previous effort, the Workflow Recognition (WR) large-scale dataset is a collection of video sequences from the real industrial manufacturing environment of a major automobile manufacturer.
In this chapter, we are addressing the issue of privacy in our modern world of Internet, Web 2.0, personalization, location based services, and ubiquitous computing. The issue is initially viewed from the perspective of user profiles, starting from existing approaches used in social networking and mobile computing applications. Emphasis is given on the separation of personal and public information and the way it can be used in Web and mobile applications. Furthermore, identifying the importance and the actual meaning of privacy in an online world is a crucial and difficult task, which has to be carried out before trying to propose ways to protect the users’ privacy.
The emergence of cloud environments has made feasible the delivery of Internet-scale services by addressing a number of challenges such as live migration, fault tolerance and quality of service. However, current approaches do not tackle key issues related to cloud storage, which are of increasing importance given the enormous amount of data being produced in today's rich digital environment (e.g. by smart phones, social networks, sensors, user generated content). In this paper we present the architecture of a scalable and flexible cloud environment addressing the challenge of providing data-intensive storage cloud services through raising the abstraction level of storage, enabling data mobility across providers, allowing computational and content-centric access to storage and deploying new data-oriented mechanisms for QoS and security guarantees. We also demonstrate the added value and effectiveness of the proposed architecture through two real-life application scenarios from the healthcare and media domains.
Man overboard incidents in a maritime vessel are serious accidents where, the efficient and rapid detection is crucial in the recovery of the victim. The severity of such accidents, urge the use of intelligent systems that are able to automatically detect a fall and provide relevant alerts. To this end the use of novel deep learning and computer vision algorithms have been tested and proved efficient in problems with similar structure. This paper presents the use of a deep learning framework for automatic detection of man overboard incidents. We investigate the use of simple RGB video streams for extracting specific properties of the scene, such as movement and saliency, and use convolutional spatiotemporal autoencoders to model the normal conditions and identify anomalies. Moreover, in this work we present a dataset that was created to train and test the efficacy of our approach.
Accurate prediction of the seawater intrusion extent is necessary for many applications, such as groundwater management or protection of coastal aquifers from water quality deterioration. However, most applications require a large number of simulations usually at the expense of prediction accuracy. In this study, the Gaussian process regression method is investigated as a potential surrogate model for the computationally expensive variable density model. Gaussian process regression is a nonparametric kernel-based probabilistic model able to handle complex relations between input and output. In this study, the extent of seawater intrusion is represented by the location of the 0.5 kg/m3 iso-chlore at the bottom of the aquifer (seawater intrusion toe). The initial position of the toe, expressed as the distance of the specific line from a number of observation points across the coastline, along with the pumping rates are the surrogate model inputs, whereas the final position of the toe constitutes the output variable set. The training sample of the surrogate model consists of 4000 variable density simulations, which differ not only in the pumping rate pattern but also in the initial concentration distribution. The Latin hypercube sampling method is used to obtain the pumping rate patterns. For comparison purposes, a number of widely used regression methods are employed, specifically regression trees and Support Vector Machine regression (linear and nonlinear). A Bayesian optimization method is applied to all the regressors, to maximize their efficiency in the prediction of seawater intrusion. The final results indicate that the Gaussian process regression method, albeit more time consuming, proved to be more efficient in terms of the mean absolute error (MAE), the root mean square error (RMSE), and the coefficient of determination (R2).
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
