Abdessamad Ait El Cadi

Université de Valenciennes et du Hainaut-Cambrésis (UVHC), ENSIAME, Faculty Member

Followers

Following

Co-authors

Public Views

Abdessamad Ait El Cadi (M). Associate Professor. With 15 years of experience in university teaching and training (~ 5000 h in Logistics, Industrial Engineering, Operational Research, Optimization, Computer Science and Mathematics), also professional experiences in the fields of Industrial Engineering, Logistics and Supply Chain. Abdessamad has develop key experiences in optimization (metaheuristics, combinatorial optimization, exact methods (PL, PLNE ...), simulation, data mining, statistical analysis, logistics, operations management, project management and information systems design. Engineer at the Ecole Nationale des Ponts et Chaussées, with a PhD from Ecole Polytechnique de Montreal, he is also member of the association "Anciens ENPC (Ecole Nationale des Ponts et Chaussés)", member of the ROADEF, member of the GDR-RO, member of ISG - Institute of Sound Management, Member of the OIQ - Order of Engineers of Quebec, previous member of APICS (American Production and Inventory Control Society, Montreal chapter). He is, also, active in scientific research with many publications, reports and conferences (10 journal articles, 20 internationals and nationals conferences, 5 seminars, 10 technical and contrat reports and one book chapter). He, also, participate to the organization of the "Forum Horizon Morocco 1998" (allowing the meeting between Moroccan companies and young people / Moroccan students in France).

less

Albert Mills

Saint Mary's University (Canada)

Steven Pinker

Harvard University

Colin Williams

The University of Sheffield

Armando Marques-Guedes

UNL - New University of Lisbon

Prof. (Dr.) Chandra K Jaggi

University of Delhi

Fabio Cuzzolin

Oxford Brookes University

Michael Morin

Université Laval

Eckart Otto

Ludwig-Maximilians-Universität München

José Francisco Pessanha

UERJ - Universidade do Estado do Rio de Janeiro / Rio de Janeiro State University

Peter Grünwald

Universiteit Leiden

Interests

Uploads

Papers by Abdessamad Ait El Cadi

Performances Modeling of Computer Vision-based CNN on Edge GPUs

ACM Transactions in Embedded Computing Systems, Mar 26, 2022

Convolutional Neural Networks (CNNs) are widely used in various ields nowadays, particularly for ... more Convolutional Neural Networks (CNNs) are widely used in various ields nowadays, particularly for computer vision applications. Edge platforms have drawn tremendous attention from academia and industry due to their ability to improve execution time and preserve privacy. However, edge platforms struggle to satisfy CNN's needs due to their computation and energy constraints. It is then challenging to ind the most eicient CNN that respects accuracy, time, energy, and memory footprint constraints for a target edge platform. Furthermore, given the size of the design space of CNNs and hardware platforms, performance evaluation of CNNs entails several challenges and eforts. Consequently, the designers need tools to quickly explore the large design space and select the CNN that ofers the best performance trade-of for a set of hardware (HW) platforms. This paper proposes a Machine Learning (ML) based modeling approach for CNN performances on edge GPU-based platforms for vision applications. We implement and compare ive (5) of the most successful ML algorithms for accurate and rapid CNN performances predictions on three (3) diferent edge GPUs in image classiication. Experimental results demonstrate the robustness and usefulness of our proposed methodology. For three of the ive ML algorithms, namely XGBoost, Random Forest, and Ridge Polynomial regression, an average error of 11%, 6%, and 8%, have been obtained for CNN inference execution time, power consumption, and memory usage, respectively.

Abdessamad Ait El Cadi

Related Authors

Uploads

Papers by Abdessamad Ait El Cadi

Log In