|
|
class JiyadHussain:
def __init__(self):
self.role = "AI Engineer"
self.focus = ["Tri-modal YOLOv5", "ByT5 Transformers", "Geometric & Photometric Priors"]
self.achieved = "Top 10 Kaggle: Social Media Extremism Detection"
self.stack = ["Python", "PyTorch", "Hugging Face", "OpenCV"]
self.off_duty = ["cooking", "football", "video games"]
def current_mission(self):
return "Bridging the gap between raw data and production-ready inference 🎯"- 🔬 Building tri-modal YOLOv5 architectures for bi-directional PCB defect detection.
- 🧩 Developed ByT5-based Akkadian→English translation pipelines with 35+ chrF++.
- 🛠️ Shipping HackPair, a real-time VS Code collab extension for hackathons.
- 🍳 Off the clock: Master of curries, football enthusiast, and wisdom seeker.
Languages & Core
Deep Learning & Vision
Deployment & Ops
| Project | What it does | Stack |
|---|---|---|
| PCB-Defect-Recognition | Tri-modal YOLOv5 processing RGB, depth, and illumination streams | YOLOv5 CV |
| Akkadian-Translation | ByT5-based pipeline for ancient text translation | PyTorch NLP |
| Extremism-Detection | NLP pipeline distinguishing extremist from benign content (Top 10 Kaggle) | Scikit-Learn |
| HackPair | Real-time code collaboration for VS Code | TypeScript |
> keep shipping · keep kaggling · keep cooking