A5004803806- Advanced Image and Video Retrieval Techniques
- Geographic Information Systems Studies
- Privacy, Security, and Data Protection
- Complex Systems and Time Series Analysis
- Mobile Crowdsensing and Crowdsourcing
- Metaheuristic Optimization Algorithms Research
- Robotics and Sensor-Based Localization
- Financial Markets and Investment Strategies
- Image Retrieval and Classification Techniques
- Data Management and Algorithms
- Video Analysis and Summarization
- Innovative Human-Technology Interaction
- Stock Market Forecasting Methods
- Image and Signal Denoising Methods
- Neural Networks and Applications
- Software System Performance and Reliability
- Opinion Dynamics and Social Influence
- Medical Image Segmentation Techniques
- Advanced Vision and Imaging
- Image and Object Detection Techniques
- Conservation Techniques and Studies
- Fire Detection and Safety Systems
- Spatial Cognition and Navigation
- Ethics and Social Impacts of AI
- 3D Surveying and Cultural Heritage
Bucharest University of Economic Studies
2016-2023
Institutul Naţional de Cercetare-Dezvoltare pentru Mecatronică si Tehnica Masurării
2020-2021
Babeș-Bolyai University
2013-2016
University of Bucharest
2016
In2 Search Interfaces Development (United Kingdom)
2012-2015
User Interface Design (Germany)
2014
Fahrenheit (Germany)
2010-2013
Abstract. In recent years, the cultural heritage (CH) sector has experienced a rapid evolution due to introduction of increasingly powerful digital technologies and ICT (Information Communication Technologies) solutions. As for many other domains, data, Artificial Intelligence (AI), Extended Reality (XR) are opening up extraordinary opportunities expanding knowledge capabilities while boosting research on innovative solutions its valorisation preservation. Being aware fundamental strategic...
The work reported in this paper aims at the development of evolutionary algorithms to register images for signature recognition purposes. We propose and develop several registration methods order obtain accurate fast algorithms. First, we introduce two variants firefly method that proved have excellent accuracy fair run times. In speed up computation, Accelerated Particle Swarm Optimization (APSO) method. resulted are significantly faster than firefly-based ones, but rates a little bit...
Purpose – Improvements and portability of technologies smart devices have enabled a rapid growth in the amount user-generated media such as photographs videos. Whilst various generation management systems exist, it still remains challenge to discover right information, for purpose. This paper aims propose an approach reverse geocoding by cross-referencing multiple geospatial data sources enable enrichment therefore better organisation searching create overall picture about places....
In an increasingly instrumented and inter-connected digital world, citizens generate vast amounts of data, much it personal, nearly all valuable. However, controlling who can collect it, limiting what they do with determining how best to protect remain deeply undecided issues. This position paper proposes a socio-technical solution based on collective awareness informed consent, whereby data collection protection norms are configured by the users themselves, using collaborative participatory...
The main aim of the reported work is to solve registration problem for recognition purposes. We introduce two new evolutionary algorithms (EA) consisting population-based search methods, followed by or combined with a local scheme. used variant Firefly algorithm conduct search, while exploration was implemented Two-Membered Evolutionary Strategy (2M-ES). Both use fitness function based on mutual information (MI) direct toward an appropriate candidate solution. A good similarity measure one...
The paper proposes a new method that combines the decorrelation and shrinkage techniques to neural network-based approaches for noise removal purposes. images are represented as sequences of equal sized blocks, each block being distorted by stationary statistical correlated noise. Some significant amount induced in blocks is removed preprocessing step, using combined with standard shrinkage-based technique. step provides initial image sequence further compressed at certain rate, component...
With the increasing abundance of technologies and smart devices, equipped with a multitude sensors for sensing environment around them, information creation consumption has now become effortless. This, in particular, is case photographs vast amounts being created shared every day. For example, at time this writing, Instagram users upload 70 million Nevertheless, it still remains challenge to discover "right" appropriate purpose. This paper describes an approach create semantic geospatial...
The work reported in the paper addresses problem of binary image registration. proposed methodology is based on an evolution strategy (ES) developed to align two images. fitness function defined terms mutual information computed between sensed and target image. We use various recombination schemes involving standard local/global convex discrete crossover procedures as well hybrid methods. hybridizations are some inner bond chromosome alleles. In final stage we evaluate similarity resulted...
Improvements and portability of technologies smart devices has enabled rapid growth in the amount user generated media such as photographs videos. Whilst various generation management systems exist it still remains a challenge to discover "right" metadata information for right purpose. This paper describes an approach extract relevant geospatial through reverse geocoding addition cross-referencing several public data sources. The extracted can be used enable enrichment with rich semantic...
In this paper is presented a new technique consisting in applying some pre-whitening and shrinkage methods followed by neural network-based supervised approach for correlated noise removal purposed.In our work the type of covariance matrix are known or can be estimated using "white wall" method.Due to data dimensionality, PCAbased compression used obtain tractable solution.The local memories neurons determined learning process based on compressed preprocessed inputs version original...
The article deals with the mechanical behaviour of a robotic arm having gyroid lattice cell based structural frame within itself.The is well know for its poriforous structure that can provide remarkable properties such as stiffness, strength and substantial energy absorption due to regular organic structure.Therefore, structures are widespread throughout various industry palette, automotive, aerospace, biomedicine industries.Due complex structure, not be obtained using classic technologies,...
The new trends in distributed computing has changed the way we do when talking about cloud infrastructures or high-performance computing. Resource virtualization technologies enabled elasticity of resource provisioning and management through easy replication virtual nodes machine migration. In order to provide high availability reliability such environments where resources are managed served form machines, specific load balancing fault strategies needed. Based on tree analysis concepts,...
In this paper, we describe the IM3I system, which provides a flexible approach to managing and publishing collections of images videos. The system is based on web services that allow automatic manual annotation, retrieval, browsing authoring multimedia. Results user evaluations, performed by professional archivists archive managers real-world deployment have confirmed easy be used delivers complete set functionalities.
The SICOLET project proposes a heterogeneous collaborative multi-agent system consisting of terrestrial vehicle equipped with serial manipulator, drone-type and an ekranoplan capable semi-autonomous operation is determined by intelligent decision control algorithms, designed for inspection intervention in areas affected various natural or anthropic disasters.Each agent cameras, so the main data acquired consist digital images.This paper focuses on fire detection being considered important...
The SIRAMAND project propose an autonomous mobile robot for collection of solid waste, specific to the urban environment, such as glass or plastic bottles, aluminium cans placed on alleys parks, streets, etc.The robot, following a certain programmable trajectory, detects collects and deposits it in container.In case damage, can be remote controlled.