Adrian Sergiu Darabant

Adrian Sergiu Darabant
Babeş-Bolyai University | UBB · Department of Computer Science

Phd Computer Science

About

57
Publications
13,348
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
226
Citations
Introduction

Publications

Publications (57)
Article
Full-text available
The human face holds a privileged position in multi-disciplinary research as it conveys much information—demographical attributes (age, race, gender, ethnicity), social signals, emotion expression, and so forth. Studies have shown that due to the distribution of ethnicity/race in training datasets, biometric algorithms suffer from “cross race effec...
Article
Full-text available
Machine learning is a branch of artificial intelligence that has gained a lot of traction in the last years due to advances in deep neural networks. These algorithms can be used to process large quantities of data, which would be impossible to handle manually. Often, the algorithms and methods needed for solving these tasks are problem dependent. W...
Conference Paper
In this paper we address the problem of hair segmentation and hair color classification in facial images using a machine learning approach based on both convolutional neural networks and classical neural networks. Hair with its color shades, shape and length represents an important feature of the human face and is used in domains like biometrics, v...
Chapter
In this paper we tackle the problem of hair analysis in unconstrained images. We propose a fully convolutional, multi-task neural network to segment the image pixels into hair, face and background classes. The network also decides if the person is bald or not. The detected hair pixels are analyzed by a color recognition module which uses color feat...
Article
Full-text available
Micro-expressions play an essential part in understanding non-verbal communication and deceit detection. They are involuntary, brief facial movements that are shown when a person is trying to conceal something. Automatic analysis of micro-expression is challenging due to their low amplitude and to their short duration (they occur as fast as 1/15 to...
Conference Paper
In this paper, we address the problem of skin tone classification in facial images, which has applications in various domains: visagisme, soft biometry and surveillance systems. We propose four skin tone classification algorithms and analyze their performance using different color spaces. The first two methods rely directly on pixel values, while t...
Conference Paper
Full-text available
Large distributed databases are split into fragments stored on far distant nodes that communicate through a communication network. Query execution requires data transfers between the processing sites of the system. In this paper we propose a solution for minimizing raw data transfers by re-arranging and replicating existing data within the constrai...
Article
Full-text available
The accurate extraction and measurement of eye features is crucial to a variety of domains, including human-computer interaction, biometry, and medical research. This paper presents a fast and accurate method for extracting multiple features around the eyes: the center of the pupil, the iris radius, and the external shape of the eye. These features...
Article
This article aims to determine the necessary characteristics of a magnetic stimulator, capable of stimulating neural tracts of the spinal cord in a healthy subject. Our previous preliminary tests had shown that the commercial clinical magnetic stimulator Magstim Rapid2 was unable to reach excitable structures within the spinal cord, and only adjace...
Conference Paper
This paper presents an eyeglasses contour extraction method that uses genetic algorithms to find the exact shape of the lenses. An efficient shape description, based on Fourier coefficients, is used to represent the shape of the eyeglasses, allowing a wide range of shapes to be represented with a small number of parameters. The proposed method does...
Article
Full-text available
This paper presents a system that automatically extracts the position of the eyeglasses and the accurate shape and size of the frame lenses in facial images. The novelty brought by this paper consists in three key contributions. The first one is an original model for representing the shape of the eyeglasses lens, using Fourier descriptors. The seco...
Article
This paper aims in interpreting the leg muscles responses recorded by electromyography during magnetic stimulation of the spinal cord by computing the electric field induced in the spinal cord and the nearby areas during this procedure. A simplified model of the spine was created and a Finite Difference Method algorithm was implemented in Matlab.
Article
Full-text available
This paper proposes an enhanced version for three clustering algorithms: hierarchical, k-means and fuzzy c-means applied in horizontal object oriented data fragmentation. The main application is focusing in distributed object oriented database (OODB) fragmentation, but the method applicability is not limited to this research area. The proposed algo...
Article
Full-text available
This paper presents a computer vision aided method for non invasive interupupillary (IPD) distance measurement. IPD is a morphological feature requirement in any oftalmological frame prescription. A good frame prescription is highly dependent nowadays on accurate IPD estimation in order for the lenses to be eye strain free. The idea is to replace t...
Article
Full-text available
This paper, introduces a new method for counting people or more generally objects that enter or exit a certain area/building or perimeter. We propose an algorithm (method) that analyzes a video sequence, detects moving objects and their moving direction and filters them according to some criteria (ex only humans). As result one obtains in and out c...
Article
The preoccupation for improving the quality of life, for persons with different handicaps, led to extended research in the area of functional stimulation. Due to its advantages compared to electrical stimulation, magnetic stimulation of the human nervous system is now a common technique in modern medicine. A difficulty of this technique is the need...
Article
Object database fragmentation (horizontal fragmentation) deals with splitting the extension of classes into subsets according to some criteria. The resulting fragments are then used either in distributed database processing or in parallel data processing in order to spread the computation power over multiple nodes or to increase data locality featu...
Conference Paper
A model is presented that predicts the electric field induced in the arm during magnetic stimulation of a peripheral nerve. The arm is represented as a homogeneous, cylindrical volume conductor. The electric field arises from two sources: the time - varying magnetic field and the accumulation of charge on the tissue - air surface. In magnetic stimu...
Article
Full-text available
In order to be able to establish the most accurate diagnostics as quick as possible, medical doctors need fast access not only to the current patient state and test results but also to its historical medical data. With the diversity of the malady symptoms today a correct diagnostic often requires a valuable time that is not always available due to...
Article
Mobile phone communication can no longer be conceived as a communication mean only, but also as a way to integrate voice services together with data services which are oriented towards large consumer groups. Together with voice services, mobile Internet represents the second most important component of the service packages offered in Romania. The a...
Article
Full-text available
Class fragmentation is an essential phase in the design of Distributed Object Oriented Databases (DOODB). Due to their semantic similarity with the purpose of database fragmentation (obtaining sets of similar objects with respect to the user applications running in the system), clustering algorithms have recently begun to be investigated in the pro...
Article
Nowadays, negotiation between a representative of a commercial enterprise and its clients is a pre-requisite for selling most of the industrial goods in large quantities. In many cases, it is the task of a mobile salesman to conduct the negotiation on behalf of the supplier. But this is not an easy task to accomplish, since the mobile agent must br...
Conference Paper
Full-text available
Mobile salesmen out in the wild and the entire commercial store with them? A while ago this would have seemed atypical. Nowadays, it has become a must have asset for any salesmen-based commercial enterprise. In the past, the Web brought the virtual store to the client's premises. While this is still enough for certain types of commerce, negotiation...
Article
Full-text available
The maximum subsequence problem is widely encountered in various digital processing systems. Given a stream of both positive and negative integers, it consists of determining the subsequence of maximal sum inside the input stream. In its two-dimensional version, the input is an array of both positive and negative integers, and the problem consists...
Article
Full-text available
Distributed relational or more recently object-oriented databases usually employ data fragmenta-tion techniques during the design phase in order to split and allocate the database entities across the nodes of the system. Most of the design algorithms are usually static and do not take into account the system evolution: data updates and addition of...
Article
Much smaller than laptops and still suitable for almost all kinds of applications, hand-held devices have the potential to rapidly become interesting tools for various daily activities. They can be successfully used in education by all participants (students, educators, administrative staff), if helpful applications are carefully designed and imple...
Article
Full-text available
A recent approach in Internet congestion control suggests collaboration between sets of streams that should share network resources and learn from each other about the state of the network. Currently such a set of collaborating streams – a macroflow – is organized on host pair basis. We propose in this paper a new method for grouping streams into m...
Conference Paper
Full-text available
In this paper we present a new approach for horizontal object oriented database fragmentation combined with fine-grained object level replication in one step. We build our fragmentation/replication method using AI probabilis- tic clustering (fuzzy clustering). Fragmentation quality evaluation is provided using an evaluator function.
Article
Full-text available
Optimal application performance in a Distributed Object Ori- ented System requires class fragmentation and the development of allocation schemes to place fragments at distributed sites so data transfer is minimal. A horizontal fragmentation approach that uses data mining clustering methods for partitioning object instances into fragments has alread...
Article
Full-text available
Horizontal fragmentation plays an important role in the design phase of Distributed Databases. Complex class relationships: associations, aggregations and complex methods, require fragmentation algorithms to take into account the new problem dimensions induced by these features of the object oriented models. We propose in this paper a new method fo...
Conference Paper
Full-text available
The Smith-Waterman algorithm is fundamental in Bioinformatics. This paper presents an FPGA-based systolic implementation of the Smith-Waterman algorithm that addresses a general case of it. A solution that improves the scalability of the design is proposed. The architecture is optimized for both speed and space, by reusing the hardware resources fo...
Conference Paper
Full-text available
Vertical and horizontal fragmentations are central issues in the design process of distributed object based systems. A good fragmentation scheme followed by an optimal allocation could greatly enhance performance in such systems, as data transfer between distributed sites is minimized. In this paper we present a horizontal fragmentation approach th...
Article
Full-text available
Class fragmentation is an important task in the design of Distributed OODBs and there are many algorithms handling it. Almost none of them deals however with the class fragmentation order details. We claim that class fragmentation order can induce severe performance penalties if not considered in the frag- mentation phase. We propose here two varia...
Article
Full-text available
Distributed Object Oriented Databases require class fragmenta- tion, performed either horizontally or vertically. Complex class relationships like aggregation and/or association are often represented as two-way refer- ences or object-links between classes. In order to obtain a good quality horizontal fragmentation, an optimal class processing order...
Article
Full-text available
Horizontal fragmentation plays an important role in the design phase of Distributed Databases. Complex class relationships: associations, aggregations and complex methods, require fragmentation algorithms to take into account the new problem dimensions induced by these features of the object oriented models. We propose in this paper a new method fo...
Article
Full-text available
As the European Union extends its boundaries the major companies have extended their presence on different markets resulting sales expansion and marketing specialization. Moreover, globalization brings a bigger impact on vital business's data because of the applications that have been developed on platforms having specific aspects by means of datab...
Article
Full-text available
Class fragmentation is an important task in the design of Distributed Object Oriented Databases (DOOD). However, fragmentation in DOOD is still at its beginnings and mostly adapted from the relational approaches. In this paper we propose an alternative approach for horizontal fragmentation of DOOD. Our method uses two different AI clustering techni...
Conference Paper
Full-text available
CREC is an original, low-cost general-purpose Reconfigurable Computer whose architecture is generated through a Hardware / Software CoDesign process. The main idea of the CREC computer is to generate the best-suited hardware architecture for the execution of each software application. The CREC Parallel Compiler parses the source code and generates...
Article
Information explosion across all areas has determined an increase in hardware requirements for application that provide data to the users. As hardware evelopment is quite susceptible to be bound after a top barrier is reached, new technologies must be developed in the software area in order to keep up with the requirements. We present here such a t...
Article
Full-text available
Class fragmentation is an essential phase in the design of Distributed Object Oriented Databases (DOODB). Horizontal and vertical fragmentation are the two commonly used fragmentation techniques. We propose here two new methods for horizontal fragmentation of objects with complex attributes. They rely on AI clustering techniques for grouping object...
Article
The preoccupation for improving the quality of life, for persons with different handicaps, led to extended research in the area of functional stimulation. Due to its advantages compared to electrical stimulation, magnetic stimulation of the human nervous system is now a common technique in modern medicine. A difficulty of this technique is the need...
Article
Full-text available
Networks expansion and Internet provide a good platform for e-learning in the idea of connecting learners with educational resources. The various systems that are already implemented consider the learning process as a remote task to gather knowledge in order to pass some exams. In the learning process evaluation represents a final step for a course...
Article
Design of modern Distributed Object Oriented Databases (DOODs) requires class fragmentation techniques. Although research has been conducted in this area, most of the developed methods are inspired from the relational fragmentation algorithms. In this paper we develop a comparative approach of two new methods for horizontal class fragmentation in a...
Article
Full-text available
Optimal application performance on a Distributed Object Based System requires class fragmentation and the development of allocation schemes to place fragments at distributed sites so data transfer is minimal. In this paper we present a horizontal fragmentation approach that uses the k-means centroid based clustering method for partitioning object i...
Article
Full-text available
Class fragmentation is an essential phase in the design of Distributed Object Oriented Databases (DOODB). Due to their semantic similarity with the purpose of database fragmentation (obtaining sets of similar objects with respect to the user applications running in the system), clustering algorithms have recently begun to be investigated in the pro...
Article
Full-text available
In order to take the appropriate decisions as quick as possible, medical doctors need fast access to various pieces of information on their pa-tients. The required information should be accurate, up-to-date, and avail-able on the spot. Even more, after finishing his/her investigation, the medical doctor should be able to immediately forward the rel...
Article
Full-text available
In order to take the appropriate decisions as quick as possible, medical doctors need fast access to various pieces of information on their patients. The required information should be accurate, up-to-date, and available on the spot. Even more, after finishing his/her investigation, the medical doctor should be able to immediately forward the relev...