- Machine Learning Engineer with a focus on Computer Vision and Deep Learning ποΈπ§
- Solid foundation in Python π , C/C++ and SQL.
- Passionate about Linear Algebra, Probability & Statistics β always learning! π€ππππ
- I enjoy building algorithms from scratch to understand their core β check out my ML Deep Dive.
- π Visit my portfolio: omartbakr.github.io
Databases & Tools:
ML/DL Stack:
Mathematical Foundations:
-
Volleyball Group Activity Recognition π (In Progress)
- Reproducing the CVPR 2016 hierarchical group-activity recognition pipeline on the Volleyball Dataset (55 videos, 4,830 clips, 8 group activities) across 8 progressively complex baselines.
- Tech: PyTorch, ResNet-50, LSTM, Hydra, TensorBoard, LMDB, Multi-GPU training.
-
Credit Card Fraud Detection π³
- Tackled extreme class imbalance (< 0.2% fraud rate) by benchmarking 11 ML models, including a custom Focal Loss implemented in PyTorch from mathematical foundations.
- Tech: PyTorch, XGBoost, LightGBM, CatBoost, Scikit-learn.
-
New York Taxi Trip Duration π
- Comprehensive feature engineering project to predict taxi trip durations.
- Tech: Python, Pandas, Scikit-learn, OSRM, Docker, Folium.
-
- SQL analysis of 700,000+ job postings exploring market trends for tech roles across countries.
- Tech: PostgreSQL, pgAdmin, Python.
-
LeetCode: 509+ problems solved (278 Medium Β· 24 Hard)
-
Codeforces: Actively solving problems, with 82 solved here
