AbdelrahmanShaker Resume

Personal Info
Name: Abdelrahman Shaker

Date of birth: 01/10/1994
Nationality: Egyptian
Residence: Masdar City, Abu Dhabi, UAE (Golden Visa Holder)
Phone number: (+971) 52-7096639
Email address: abdelrahman.m.shaker@gmail.com
GitHub website: https://amshaker.github.io/
LinkedIn profile: www.linkedin.com/in/abdel-rahman-shaker-74274b9b
Profile Summary
Have a mixed experience between conducting pure academic research and contributing to international
products in the industry. My research interest is Computer Vision, specially designing accurate, lightweight,
and memory-efficient 2D and 3D architectures for edge devices. Seeking for research internship or a
graduate fellowship program related to the field of Computer Vision/AI during my Ph.D. journey.
Education
Ph.D. in Computer Vision
MBZUAI 08/2021 – Present
- CGPA is 3.95/4.00 (A).
Ain-shams University Master degree in Computer Sciences 03/2018 – 08/2020

- Master Courses (2017 / 2018).
- CGPA is 3.71/4.00 (A-).
Ain-shams University Bachelor degree in Computer Sciences 09/2012 – 08/2016
- Excellent with honors grade.
- Dept Rank: 1st (out of 149 students).
Work Experience
Valeo Egypt Machine Learning / SW Engineer 11/2019 – 7/2021
Ain-shams University Teaching / Lecturer Assistant 03/2018 – 7/2021
Valeo Egypt Deep Learning Researcher 01/2019 – 04/2019

Internship at CDV-R&D Excellence team
Mercedes Benz-GAS Software Engineer 09/2016 – 01/2017
ABB Software Engineer 08/2016 – 09/2016

Internship
Selected Publications
• SwiftFormer: Efficient Additive Attention for Transformer-based Real-time Mobile Vision
Applications, Accepted in ICCV conference, 2023.
• UNETR++: Delving into Efficient and Accurate 3D Medical Image Segmentation, Submitted
to IEEE TMI.
• EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision
Applications, Published in ECCVW, 2022.
• Generalization of Convolutional Neural Networks for ECG Classification Using Generative
Adversarial Networks, Published in IEEE Access Journal, 2020.
Projects
• XrayGPT: Conversational medical vision-language model, capable of analyzing chest radiographs

and providing insightful answers to open-ended questions based on the given x-ray [project].
• Driver Monitoring System (DMS): Integral contributor to DMS Daimler project, focusing on driver
recognition and alertness within an integrity OS framework. I focused on integrating components from
ECP, Vision, and Algorithms teams, aligning with customer requirements (SRS). Demonstrated
proficiency in cross-functional collaboration, intricate integration, and advanced hardware utilization.
• Insta-YOLO: A novel one-stage end-to-end DL model for real-time instance segmentation, the box
regression of YOLOv3 is replaced by a polygon regression in the localization head in addition to
proposing new localization loss. The model is 1.75x faster than YOLACT with comparable accuracy.
• Seat Occupancy Detection: Detect & classify the car seats in order to know how to open the airbag
for more safety. I increased the Average Precision (AP) by 4.0% for different versions of YOLO and
CenterNet by proposing a different data augmentation method that is based on mixing image
normalization techniques.
• Object Recognition using Radar: Using TI radar sensors interior of the car, I introduced a
prototype of classifying the objects from the heatmaps.
• Occupancy Grid Mapping: Autonomous driving requires detailed knowledge about the
environment. A common approach to accomplish this task is to use occupancy grid maps (OGM). We
built our own OGM based on LIDAR data. I reduced the time from 3.18 MS to 1.28 MS by updating
only the polygon of the affected points instead of updating the whole convex hull.
• Arrhythmia Classification: In my master thesis, I proposed a generalization method using GANs
based on different deep learning architectures (CNN, LSTM, and CNN+LSTM) that improved the overall
accuracy and precision of the public MIT-BIH benchmark.
Teaching History
• Teaching courses: Machine Learning, Deep Learning, and Computer Vision.
• Mentor in Kaggle competitions for deep learning projects at Ain Shams University.
• Supervisor for several Machine Learning & Computer Vision graduation projects.
Technical Skills
Programming Languages: Python and C++. Frameworks: PyTorch and Tensorflow.
Concepts: OOP, Data structures, Analysis and design algorithms, and agile methodologies.
Awards & Honor Certificates
• Ranked 1st in Deep Learning course during Ph.D. at MBZUAI.
• First place in Ibtiecar competition 2017 (Hold by the Egyptian Ministry of Communications)
o Ibtiecar is a competition held every year for excellent graduation projects for all engineering and
computer sciences colleges in Egypt. We had achieved first place in the Data Science track.
• Coursera certificate for completing the following specializations: -
o TensorFlow in Practice specialization from deeplearning.ai
o Deep Learning specialization from deeplearning.ai
o Machine Learning specialization from the University of Washington
• Ranked in the top 3 in the Algorithms and Operating System courses.
Academic Supervisor
• Prof. Dr. Fahad Khan (fahad.khan@mbzuai.ac.ae)
Languages
• Arabic: Native or bilingual proficiency.
• English: Professional working proficiency.

AbdelrahmanShaker Resume

Uploaded by

Copyright:

Available Formats

AbdelrahmanShaker Resume

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AbdelrahmanShaker Resume

Uploaded by

Copyright:

Available Formats

Personal Info

Name: Abdelrahman Shaker

Ain-shams University Master degree in Computer Sciences 03/2018 – 08/2020

Ain-shams University Teaching / Lecturer Assistant 03/2018 – 7/2021

Valeo Egypt Deep Learning Researcher 01/2019 – 04/2019

Mercedes Benz-GAS Software Engineer 09/2016 – 01/2017

ABB Software Engineer 08/2016 – 09/2016

• XrayGPT: Conversational medical vision-language model, capable of analyzing chest radiographs

You might also like