Categories

Tackling Continual Learning: Our Journey in the CVPR 2021 Challenge

Check the code 

In the rapidly evolving field of AI, one of the biggest hurdles is teaching machines to learn sequentially, like humans do, without forgetting previously acquired knowledge. This is the core challenge of Continual Learning (CL). In 2021, we participated in the...

Autonomous Lane Following in Duckietown: Our DAgger Approach

Check the code 

Ever wondered how miniature autonomous cars learn to navigate? At Duckietown, a platform designed for education and research in robotics, we tackled the challenge of lane following and vehicle avoidance using an advanced imitation learning technique. This project, which was also...

Avito Demand Prediction Kaggle competition

Check the code 

Avito launched a competition on Kaggle challenging users to predict Avito to predict demand for an online advertisement based on its full description (title, description, images, etc.), its context (geographically where it was posted, similar ads already posted) and historical demand for similar ads...

Multi-lingual Bot Framework V4

Check the code  During my work at Microsoft Research Lab in Cairo , i was lucky to work on bot framework project to support multiple languages.

Learning Speech Recognition from Tf Speech Commands

Check the code 

At the end of 2017, Google Launched a competition on Kaggle using its dataset Speech command. In this competition we were challenged to predict simple commands from input user speech command. Each utterance is around 2 seconds.

This competition attracted me to gain more...

Model Generation Library

Check the code 

During my work at Valeo , due to an important project request i was able to gain more experience in both MATLAB ( used before in a previous computer vision project ) and Simulink models.

Action Classification from Still Images

As a final project in the Computer Vision course at my college with Prof Marwan Torki, we used Pascal Voc 2010 Action Classification. This project excited me to learn more about machine learning and computer vision in specific. Since that project machine learning became a hobby to me and Kaggle platform made my life easier to learn new techniques...

Windows Phone Development

At my second year at Alexandria University I was attracted to the field of mobile application development. At that time I found Microsoft windows phone 7.1 SDK to be suitable to start mobile development with C# in a new promising mobile platform.

Allemny Initiative Website

Allemny is an Arabic word that means teach me, this initiative mainly aims to encourage self-Learning and to keep teaching techniques up to date by using new education tactics to cope with the modern technologies. We are creating short videos that explain engineering topics and software in an easy way, so that everybody can understand our data.

In that project,...

Avito Demand Prediction Kaggle competition

Check the code 

Avito launched a competition on Kaggle challenging users to predict Avito to predict demand for an online advertisement based on its full description (title, description, images, etc.), its context (geographically where it was posted, similar ads already posted) and historical demand for similar ads...

AI Challenger English Chinese Machine Translation

Ai Challenger a new Chinese platform for ai challenges, their first contest was related to machine translation system and I wanted to try my techniques in NMT systems on a system that I have no clue about its target Chinese language.

In that contest I participated as Marb and got 25.50 bleu score on their evaluation .

Learning Speech Recognition from Tf Speech Commands

Check the code 

At the end of 2017, Google Launched a competition on Kaggle using its dataset Speech command. In this competition we were challenged to predict simple commands from input user speech command. Each utterance is around 2 seconds.

This competition attracted me to gain more...

Building Reservation System

During Summer 2013, Faculty of Engineering Alexandria University announced a competition to make an online room reservation system with automated filling of rooms based on available courses along with their capacity and required equipment in the room, I won the first prize of this contest.

Participating in ACM Sigmod 2013 EGN Team

ACM Sigmod Programming Contest is a programming contest launched 2013 , mainly focused on implementing a program used to match large documents indexed by the tool in runtime.

This was my first worldwide online contest and it was an exciting experience for me and for my colleagues with EGN Team.

Working at Raisa

I was able to get an offer from Raisa Egypt , a startup specialized in forecasting oil and gas from land geological features.

Working at Microsoft Research Lab

I have always wanted to work in the field of machine learning and during my work at Valeo , I was participating in kaggle for learning purposes. Moving to Microsoft Lab helped me gain more industrial and research experience in the field.

Working at Valeo

After graduating from Faculty of Engineering Alexandria university, I was lucky to get a recommendation at Valeo Automotive company, for 2 years and 3 months at Valeo I gained a broad knowledge in software development.

Bkam Internship as Software Engineer

Bkam was a fast-growing startup specialized in online price comparison between available online and on-site products from several stores and get you the store link with the best price for a specific item.

At my second year I was lucky to get a recommendation from my TA Ahmed ElSharkasy to join them.

GROOD - GRadient-aware Out-Of-Distribution detection

Out-of-Distribution (OOD) detection, the task of identifying inputs that a model has not been trained on, is fundamental for the safe and reliable deployment of deep learning models in real-world applications like autonomous driving and healthcare. Models that perform well on familiar, in-distribution...

BACS - Tackling Background Ambiguity in Continual Semantic Segmentation

Semantic Segmentation, the task of assigning a class label to every pixel in an image, is fundamental for detailed scene understanding, especially in applications like autonomous driving and robotics. However, creating these pixel-perfect annotations is laborious. Furthermore, real-world systems often need to learn...

Identifying Critical Neurons in ANN Architectures using Mixed Integer Programming

In this research, we attempt to understand the neural model architecture by computing an importance score to neurons. The computed importance score can be used to prune the model or to understand which features are more meaningful to the trained ANN (artificial neural...

Gender aware spoken language translation Arabic To English

During my work at Microsoft Research Lab in Cairo, we were brainstorming for research projects related to our work Skype Translator for the upcoming summer internship.I got the idea of making a machine translation system that keeps the lost gender information while translating from Arabic To English.

Synthetic Spoken Data for Neural Machine Translation

Download the paper

During my work at Microsoft Research Lab in Cairo , my first task was to make a Levantine Arabic dialect to English NMT system, but the parallel data available for dialectical Levantine was not enough to train a decent model for Skype translator.

Dev on Docker, Deploy on Singularity The MLOps Workflow You've Been Missing

You’ve perfected your machine learning model in a Docker container on your local machine. But how do you run it on a secure, shared high-performance computing (HPC) cluster? The answer is Singularity. Let’s walk through the process.

Introduction to Mixed Integer Programming

Mixed-Integer programming are used to solve optimization problems with discrete decision variables. Hence, its feasible region is a set of disconnected integer points and gradient based algorithms cannot be directly applied.

Learning Attentional LSTM model

Out-of-Distribution Detection in Vision-Language Models: A Survey

Vision-Language Models (VLMs) like CLIP have dramatically shifted the landscape of visual understanding. Trained on internet-scale image-text pairs, these models demonstrate remarkable zero-shot generalization, describing objects they have never explicitly seen during training. Yet this generalization comes with an underappreciated fragility: when deployed in the real world, VLMs routinely encounter inputs that bear no resemblance to anything...

Reasoning's Razor: When Thinking More Makes Safety Worse

Large Reasoning Models (LRMs) like DeepSeek-R1 and QwQ-32B have become remarkably capable at solving complex problems through extended chain-of-thought. The natural instinct is to apply this power to safety-critical tasks: detecting harmful content, catching hallucinations, flagging policy violations. More reasoning = more accuracy = safer AI, right?

A new paper challenges that intuition head-on. “Reasoning’s Razor” <a...

How Can LoRA parameters improve the detection of Near-OOD data?

We’ve all come to love Low-Rank Adaptation (LoRA) for making it practical to fine-tune massive Large Language Models (LLMs) on our own data. The standard practice is simple: you inject small, trainable matrices into the model, fine-tune only them, and then, for deployment, you merge these new weights back into the original model to avoid any inference...

Weight Space Learning Treating Neural Network Weights as Data

In the world of machine learning, we often think of data as the primary source of information. But what if we started looking at the models themselves as a rich source of data? This is the core idea behind weight space learning, a fascinating and rapidly developing field of AI research. The real question in this post...

Decoding LLM Hallucinations An In-Depth Survey Summary

The rapid advancement of Large Language Models (LLMs) has brought transformative capabilities, yet their tendency to “hallucinate”—generating outputs that are nonsensical, factually incorrect, or unfaithful to provided context—poses significant risks to their reliability, especially in information-critical applications . A comprehensive survey by Huang (Huang et al., 2025) systematically explores this phenomenon, offering a detailed taxonomy, analyzing root causes,...

Topology of Out-of-Distribution Examples in Deep Neural Networks

As deep neural networks (DNNs) become more common, concerns about their robustness, particularly when facing unfamiliar inputs, are growing. These models often exhibit overconfidence when making incorrect predictions on out-of-distribution (OOD) examples. This...

Out-of-Distribution Detection Ensuring AI Robustness

Deep neural networks can solve various complex tasks and achieve state-of-the-art results in multiple domains such as image classification, speech recognition, machine translation, robotics, and control. However, due to the distributional shift between collected training data and actual test data, The trained neural network has a difference between the network’s performance on the training and unseen real...

GAIA Gradient-Based Attribution for OOD Detection

Deep neural networks (DNNs) have shown incredible accuracy across numerous applications. However, their inability to handle out-of-distribution (OOD) samples can lead to unpredictable and potentially unsafe behavior. This post explores the recent paper on the Gradient Abnormality Inspection and Aggregation (GAIA)(Chen et al., 2023) framework, which introduces an innovative approach to enhance OOD detection.

Gradient-aware...

Survey on Uncertainty Estimation in Deep Learning

A distinction between aleatoric and epistemic uncertainties is proposed in the domain of medical decision-making (Senge et al., 2014). Their paper explained that aleatoric and epistemic uncertainties are not distinguished in Bayesian inference. Moreover, the expectation over the model with respect to the posterior is used to get our prediction leading to an averaged epistemic...

Predicting Neural Network Accuracy from Weights

Check the code 

In this paper (Unterthiner et al., 2020) showed empirically that we can predict the generalization gap of a neural network by only looking at its weights. In this work, they released a dataset of 120k convolutional neural networks trained...

Predicting The Generalization Gap In Deep Networks With Margin Distributions

Check the code 

In this paper (Jiang et al., 2018), they discuss a method that can predict the generalization gap from trained deep neural networks. The authors used marginal distribution information from input training set as a feature vector used by an...

Explaining PackNet Adding Multiple Tasks to a Single Network by Iterative Pruning

Check the code 

In this paper (Mallya & Lazebnik, 2017), they discuss a method for adding and supporting multiple tasks in a single architecture without having to worry about catastrophic forgetting . They show in this paper that three fine-grained classification tasks can be added...

Stabilizing GAN training With Random Projections

Check the code 

In this paper (Neyshabur et al., 2017), They introduced a framework to stabilize the GAN training by using multiple projections with fixed filters of each input image to a different discriminator. Training GAN models is unstable in high dimensional space and some...

Empowering the Next Generation My Journey with MISE

In the ever-evolving landscape of artificial intelligence, there’s a pressing need to cultivate diverse talent and foster inclusivity in STEM fields. One initiative that has personally enriched my journey and allowed me to contribute meaningfully to this cause is MISE (Machine Learning in Science and Engineering), a transformative program dedicated to equipping high school...