Marco Vernier

Situational awareness in smart environments: socio-mobile and sensor data fusion for emergency response to disasters

Abstract Traditional situational awareness services in disaster management are mainly focused on the institutional warning response and not fully exploit the active participation of citizens involved. This paper presents an advanced system for emergency management (ASyEM) which fuses the potentiality offered by mobile social data and bottom-up communication with smart sensors. The proposed architecture model is organized into four different layers: (1) sensor, (2) local transmission, (3) network and (4) management. ASyEM is able to capture and aggregate two different kind of data: (a) user generated content produced by citizens during or immediately after the disaster and shared online through socio-mobile applications and (b) data acquired by smart sensors distributed on the environment (i.e., intelligent cameras, microphones, acoustic arrays, etc.). Data are selected, analysed, processed and integrated in order to increase the reliability and the efficiency of whole situational awareness services, localize the critical areas and obtain in this way some relevant information for emergency response and completion of search and rescue operations.

Remote feature learning for mobile re-identification

This work introduces a novel method for person re-identification using embedded smart cameras. State-of-the-art methods address the re-identification problem using global and local features, metric learning and feature transformation algorithms. Such methods require advanced systems with high computational capabilities. Nowadays, there is a growing interest in security applications using embedded cameras. Motivated by this we propose to study a new system that addresses the challenges posed by the reidentification problem using devices (e.g. smartphones, etc.) that have limited resources. In this work we introduce a novel client-server system that exploits a feature learning method to achieve a two-fold objective: (i) maximize the re-identification performance over time and (ii) reduce the required computational costs. In the training phase, state-of-the-art features are selected considering both the device capabilities and re-identification performance. During the detection phase, the re-identification performance are maximized by selecting the best features for a given input image. To demonstrate the performance of the proposed method we conduct the experiments using different mobile devices. Statistics about feature extraction and feature matching are presented together with re-identification results.

A Smart Visual Information Tool for Situational Awareness

In the last years, social media have grown in popularity with millions of users that everyday produce and share online digital content. This practice reveals to be particularly useful in extra-ordinary context, such as during a disaster, when the data posted by people can be integrated with traditional emergency management tools and used for event detection and hyperlocal situational awareness. In this contribution, we present SVISAT, an innovative visualization system for Twitter data mining, expressly conceived for signaling in real time a given event through the uploading and sharing of visual information (i.e., photos). Using geodata, it allows to display on a map the wide area where the event is happening, showing at the same time the most popular hashtags adopted by people to spread the tweets and the most relevant images/photos which describe the event

Selection of Temporal Features for Event Detection in Smart Security

Scene understanding in smart surveillance and security is one of the major fields of investigation in computer vision research and industry. The ability of a system to automatically analyze and learn the events that occur within a scene (e.g., a running person, a parking car) is conditioned by several complex aspects such as feature extraction, tracking and recognition. One of the most important aspects in the event learning process is the detection of the time interval in which an event occurs (i.e., when it starts and ends). The present paper is focused on the learning of temporal correlated events. In particular, a formalized description of the features associated with each event and the linked strategy to define the event time-line are provided. The paper also reports preliminary tests carried out on videos related to a reference outdoor environment which validate the proposed strategy.