Ariyan Zarei

Ariyan Zarei

Ariyan Zarei

Education

Undergreaduate

University: Shahid Beheshti University, Tehran, Iran.
Field of Study: Computer Science
Faculty: Faculty of Mathematical Sciences
Department: Computer Science Department
GPA: 3.95 / 4.0
Duration: Fall 2013-Winter 2017
Courses: [Download]

Diploma

High School: Dehkhoda, Karaj, Iran
Field Of Study: Mathematics
GPA: 19.61 / 20
Duration: Fall 2009-Fall 2013

Publications & Lectures

Journal Papers

Improve CAPTCHA’S Security Using GAUSSIAN BLUR Filter

Journal: Signal & Image Processing: An International Journal (SIPIJ)
Journal Type: International Journal
Published Date: Octobor 2014
Volume & Number: Volume 5, and Number 5
Full Text: [PDF]
Journal Website: SIPIJ

Conference Papers

A Feature Vector for Optical Character Recognition (Accepted)

Conference: Second International Conference on Frontiers of Image Processing
Conference Type: International Conference
Conference Date: 9~11 March 2018
Location: Barcelona, Spain
Presentation Type: To be Specified
Conference Website: ICFIP 2018
Full Text: -

The Effect of Applying Gaussian Blur Filter on CAPTCHA’s Security

Conference: First International Conference On Signal And Image Processing (SIGL 2014)
Conference Type: International Conference
Conference Date: 13~14 September 2014
Location: Banglore, India
Presentation Type: Oral Presentation
Conference Website: SIGL
Full Text: [PDF]

Computer Controlling System Using Text Messages

Conference: International Conference On Computers, Information Technologies and Digital Medias (CITADIM 2013)
Conference Type: International Conference
Conference Date: March 2014
Location: Tehran, Iran
Presentation Type: Oral Presentation
Full Text: [PDF-Persian]

Lectures

Image Enhancement

Place/Occasion: Shahid Beheshti University, Image Processing Course
Date and Time: Saturday, Esfand 21st - March 11th 2017, second semester of 1395-1396
About the Lecture: We reviewed the Term "Image Enhancement" and it's techniques in this lecture as a part of our course, Image Processing by DR. Alireza Tavakoli Targhi. We also provided a short explanation on Fuzzy Sets/Logic and it's application on Image Enhancement.
Lecture Slide: [PDF]
Lecture File: [RAR File]

An Introduction to Neural Networks and Deep Learning

Place/Occasion: Shahid Beheshti University, Image Processing Course
Date and Time: Saturday, Ordibehesht 9th - April 29th 2017, second semester of 1395-1396
About the Lecture: The presentation was a part of Image Processing course by Dr. Alireza Tavakoli Targhi which was presented by me. It included basic concepts of artificial neural networks and also a brief introduction to deep learning, more specificaly on the topic of CNNs and RNNs.
Lecture Slide: [PDF]
Lecture File: [Matlab m File]

Achievements and Memberships

Memberships

  1. Secretory of Scientific Association of Computer Science Students, Shahid Beheshti University (2015-2016).
  2. Member of Iran’s Society of Image Processing & Computer Vision (2015-now).
  3. Member of Youth Researchers and Elite Club, Islamic Azad University (2014-2016).
  4. Editorial Board of “Pardazeh”, The Scientific Journal of Computer Science Association, Shahid Beheshti University (2013-2014).

WorldSkills

  1. Attendanded in the Training Camp of the 43rd WorldSkills Competitions.
  2. Ranked Third and achieved Bronze medal in the 14th national skill competitions in the IT-Software solution for business field.
  3. Ranked Fifth and achieved Honoring diploma in the 13th national skill competition in the IT-Software solution for business field.
  4. Ranked First in the all levels of the Provincial and county levels in the IT-Software solution for business field.

Khwarazmi

  1. Chosen Project in the 14th Youth Khwarizmi Competition Provincial Level and acceded to the national level. Project : Computer Controlling Using SMS
  2. Chosen Project in the 13th Youth Khwarizmi Competition County Level. Project : Advanced Survilance on Automobile Gates

Competitions

  1. Ranked first in the "Gomanesh" AI/NLP competition, a 20 question competition between bots.
  2. Took Part in the national level of 13th “laboratory and Workshop Competitions” in the computer field.
  3. Ranked First in the Provincial and County level of 13th “laboratory and Workshop Competitions” in the computer field.

Skills

Skills

Keras-Deep Learning Library

OpenCv

Python

Matlab

C & C++

Kinect Programming using C#

Object Oriented Programming - C#.Net & Java

SQL Server Database Design & Develop

Maple

Programming with R for Data Mining purposes

Licenses

Windows forms applications with C#.Net

Database Design and Develop with SQL Server

ADO.Net in Entity Framework with C#.Net

Fundamental of programming with C#.Net

Visual Basic Programming course (Fundamental)

Visual Basic Programming course (Professional)

Projects

A Deep Learning Approach for Detecting Motorcycle Riders with No Helmet on (Bachelor Thesis)

Nowadays motorcycle drivers comprise a large portion of casulties related to the traffic accidents. By wearing appropriate helmets, they can prevent catastrophic consequences. Thus, we have been developing a system which is capable of detecting motorcycle drivers who are not wearing helmets. We want to employ VGG-16 to detect helmets and motorcycles in the input images. We are mainly using Keras and Tensorflow libraries. Demo: [Download]

Partners : Dr. Alireza Tavakoli Targhi


Pedestrian Detection using SVM classifier and HOG feature

We managed to implement a Pedestrian Detection system in Matlab. The first step towards the detection is to train the system which in our case it was done by means of SVM classifier and positive/negative image set provided by Cornel University. Then it was time to move forward to have a complex pedestrian detector system which detects and highlights the pedestrians with various sizes in a large and detailed image, containing other objects. you can find the code files [here] and also you can read the project report [here].

Partners : Dr. Alireza Tavakoli Targhi


Visual Cryptography

An Application designed by c# programming language. This App is a tool for Visual Cryptography Technique. It will Capture a picture from a camera, then it will Dithere the image using an Extension of Floyed-Steinburg algorithm and using Otsu Treshoulding. The App will produce 2 Shares from the dithered Image.

Partners : Dr.Katanforoush


Using Kinect gesture recognition capability to treat patients with psychological disorders

In this project, which it was a joint project in the fields of computer vision, machine learning, and cognitive science, we used Microsoft Kinect to develope a gesture recognition app. our purpose was to cure kids with special mental disorders by asking them to imitate the movements which the app shows them.

Partners : Dr.Katanforoush, Dr.Nejati


Digit Recognition with SVM classifier and HOG feature

This project deals with the recognition of english digits using SVM classifier and HOG feature vector. In order to find the best SVM classifier, we searched for the best HOG feature vector. It seemed that having HOG feature vectors produced by a 2 by 2 window, yields the best results and is accurate with the rate 90%. you can find the code files [here], results [here] and also you can read the project report [here].

Partners : Dr. Alireza Tavakoli Targhi


Implementing the Edge Detection Methods in Matlab

As it is mentioned in the title, we implemented the Sobel and Canny edge Detection methods in matlab with various parameters. this project was a project of the image processing course in my undergrad. you can find the code files [here] and also you can read the project report [here] .

Partners : Dr. Alireza Tavakoli Targhi