Monday 20 November 2017
Chairperson of the Dept. of CCE at AUST Participates in the Complex Adaptive Systems Conference CAS 2017, Chicago, USA

Dr. Roger Achkar, Chairperson of the Department of Computer & Communications Engineering at AUST participated in the Complex Adaptive Systems Conference CAS 2017, which was on Engineering Cyber Physical Systems. It was held in Chicago, USA, from October 30 till November 1st 2017.

Dr. Achkar represented AUST at the conference, and presented two article papers: The first paper is entitled “LPR CNN Cascade and Adaptive Deskewing”. The paper presents the work of an MS project done at AUST- Ashrafieh. It is co-authored by Mr. Georges Bou Kassm.

The second paper is entitled “ERP Neural Network Inventory Control”. The paper presents the work of an MS project done at AUST- Ashrafieh. It is written by Mr. Michel Owayjan, Chairman of the Mechatronic Department at AUST, and co-authored by Mr. Jad Farhat.

CAS Conferences was known previously as ANNIE Conferences; the ANNIE Conferences, From 1991 – 2010, provided an international forum for researchers who were interested in engineering applications of Fuzzy Logic and Systems, Neural Networks and Evolutionary Programming in St. Louis, Missouri, USA. Complex Adaptive Systems: In 2011 the ANNIE Conference series was concluded and a new series began under a broader focus, entitled Complex Adaptive Systems. The conference and publication structure was designed to enable a different theme and location each year. The first year of the conference was a great success with over 100 participants from around the world that met in Chicago, Illinois.  Since then the conference has moved to Washington D.C. in 2012, Baltimore, MD in 2013, Philadelphia, PA in 2014, San Jose, CA in 2015 and Los Angeles, CA in 2016 and has continued to bring together experts in academia, industry and government to push the boundaries of knowledge of Complex Systems to the next level.
These two article papers were published in the Procedia of Computer Science, Elsevier, Volume 114, November 2017.

LPR CNN Cascade and Adaptive Deskewing- Abstract — License Plate Recognition (LPR) is a well-known image processing technology. The goal of this paper is to achieve the same or better accuracy than previous known algorithms while attaining a higher speed of processing and a modular yet simple approach. The engine is mainly designed to work on Lebanese license plates but can easily be trained for others. The results below are obtained by using a deep convolutional neural network cascade for classification (CNN cascade), a CNN with partially connected deep layers for deskewing and a neural network optimized by neuro-evolution for OCR. This resulted in a modular LPR solution that surpasses the conventional solution in terms of speed and accuracy, a deskew module that can straighten double lined plates with far better accuracy than its image processing counterpart and an OCR module that’s optimized for the best speed and accuracy achievable even predicting characters that are unreadable by conventional solutions. On top of that, the whole solution is GPU (Graphical Processing Unit) enabled making it scalable for a large network of cameras with a central processing unit.

ERP Neural Network Inventory Control - Abstract — Enterprise Resource Planning (ERP) is a system of integrated applications used to have full insight over the resources of an enterprise in terms of goods, employees and customers. On the other hand, artificial neural networks are becoming a necessity in applications that require artificial intelligence. A marriage between these two concepts would yield a system capable of storing and displaying dashboards of data, and simultaneously make computed expectations that can determine the future plans of an enterprise. Many have researched different applications in which a neural network can be used in order to achieve such a system. This paper demonstrates the study and simulation of a system that can give a prediction of the goods needed for an enterprise’s inventory depending on the past history of this enterprise sale with respect to the events occurring at different time periods. The system is built using C# and using examples from a real trading cooperation history in the learning process. It was tested using fictional simulations and produced acceptable results.

Official AUST Website
All rights reserved. © 2018