Rezq is a passionate software engineer, specializes in developing efficient machine learning models. Rezq worked on many academic projects using Python, Matlab, and C+. He is proficient in Data Structures and Algorithms, Object Oriented Programming,
The System depends mainly on Two Stages, First Stage: Take an input X-Ray image of the upper extremity part of the body then enhance the image to show the details in the bone better, then take the enhanced image to a Deep Learning model to classify its type into one of 7 types whether it is [Shoulder, Humerus, Elbow, Forearm, Wrist, Hand or Finger].
Second Stage: We have 7 different Deep Learning models (one for each Bone), So depending on the bone type that is predicted in the first stage we take the enhanced image in the first stage to its corresponding model (based on the predicted bone type) to predict its condition whether it be a normal or abnormal Bone.
Image Classification based on the contextual data of the image. Firstly we applied several data preprocessing, for example: cropping the cars using it’s boundary boxes then applying many data augmentation techniques: Flipping with Fixed degrees,
Noise addition and Applying Grayscale to the images. After that, we train different models: using a CNN architecture and Pre-Trained VGG Model.
This project aims to learn how to detect a human face from a recorded or-real video (live cam) and recognize the gender of the human and his/her facial expression. Using multiple feature extraction techniques (HAAR cascade, HoG and HoF Descriptors) and for classification phase: CNN and SVM.
Make a computer play game based on image processing techniques, without using any keyboard keys. Creating a set of frames for each Move/Action. First the computer analyze the scene to detect and recognize the location of (Mario, Coins, Enemies, Blocks, Castle) then the computer starts to make a strategy to pass the level given these information, with the goal to collect all coins detected in the level and finally go to the castle to pass, if there is an enemy in his way he should kill it first by
using the attack action.
Encrypt images by applying LFSR algorithm and compress them by using priority queue data structure and vice versa ( Added more Powerful features to Strength the Power of the encryption by encrypting the text key and password then converting them to their corresponding ASCII binary representation).
Designing small compiler for Assembly Language with combining instructions like (INC, ADD, SUB, MOV, MOVZX, MUL, DIV, call dumpregs, call write int ) using EAX, EBX, ECX, EDX, ESI, EDI, ESP, EBP registers.
Python, C, C+, Matlab, C Sharp
Pandas, Keras, Tensorflow
Bachelor’s Degree in Computer Science, Ain Shams University (2019)
Achievements and Awards:
Rezq is now available for hireHire Rezq