I am a Computer Science undergraduate at Namal University Mianwali (CGPA 3.10/4.0, Merit-Based Scholarship Recipient), committed to building strong foundations in Computer Vision, Deep Learning, and Machine Learning.
I am not yet an engineer. I am a student actively learning, building real projects, and growing toward that goal. My approach is hands-on: I learn by implementing papers, breaking models, fixing bugs, and shipping working systems.
Currently:
- Working on: Adversarial Privacy Mask (custom loss function research)
- Learning: MIT 6.S191 Deep Learning, Andrew Ng ML Specialization
- Building: Distributed Cloud Application on Microsoft Azure
- Exploring: Adversarial ML, Computer Vision papers, Model Optimization
Aspiring to become:
- Computer Vision Engineer
- Machine Learning Engineer
- AI Research Engineer
Open to:
- Software Engineering, Machine Learning, Computer Vision, or Backend Internships
- Locations: Remote or Pakistan-based
|
Status: Ongoing Research A team project where I am co-developing a custom loss function integrated with PGD attacks for adversarial perturbation generation. The approach empirically outperforms standard FGSM and PGD baselines on facial recognition models. Currently scaling validation from LFW to CASIA-WebFace using PyTorch. Tech: Python, PyTorch, Adversarial ML, CASIA-WebFace |
Status: Completed End-to-end computer vision system that detects and counts date palm trees from aerial imagery using YOLOv11 deep learning. Integrated BoT-SORT multi-object tracking for video frame consistency. Built Flask REST API with React Native mobile frontend. Tech: Python, YOLOv11, BoT-SORT, OpenCV, Flask, React Native |
|
Status: Ongoing A team project where I architected a scalable distributed web application using MongoDB sharding and replica sets. Deployed on Microsoft Azure App Service with JWT authentication and Role-Based Access Control across three distinct user roles. Tech: Node.js, Express, MongoDB, Microsoft Azure, JWT |
Status: Completed Designed a secure enterprise network in Cisco Packet Tracer with DNS-based load balancing, AAA security, VLAN segmentation across four departments, and four Layer-2 protection mechanisms including DHCP Snooping, DAI, Port Security, and NAT. Tech: Cisco Packet Tracer, Cisco IOS, Network Security, AAA |
|
Status: Completed A statistical machine learning model that forecasts precipitation events using conditional probability and Naive Bayes classification on four atmospheric variables: temperature, humidity, air pressure, and wind speed. Includes complete data preprocessing and visualization pipeline. Tech: Python, Naive Bayes, NumPy, Pandas, SciPy, Matplotlib |
Status: Completed Full generative art web application built in pure HTML, CSS, and JavaScript with no frameworks or dependencies. Features four drawing modes, undo/redo history, PNG export, local gallery storage, and 60 FPS performance. Deployed on Microsoft Azure. Tech: HTML, CSS, JavaScript, Microsoft Azure |
I am committed to continuous learning through high-quality open educational resources:
- MIT 6.S191 — Introduction to Deep Learning (CNN, RNN, modern architectures)
- Andrew Ng Machine Learning Specialization — Supervised Learning, Neural Networks
- Ultralytics YOLOv8/v11 Official Tutorials — Object detection, model training, custom datasets
- Roboflow Computer Vision Tutorials — Annotation, augmentation, deployment (Verified Student)
- Murtaza's Workshop OpenCV Tutorials — Image processing, real-time CV applications
- LinkedIn Learning — Test-Driven Development in C++, Design Patterns (Creational, Structural, Behavioral)
- Merit-Based Scholarship Recipient — Namal University Mianwali, awarded for academic excellence
- Project Leadership Award — Recognized for outstanding teamwork and project management
- Head Boy — Fazaia Inter College, Murree (led 300+ students, organized 15+ college events)
- Roboflow Verified Student — Granted access to professional computer vision tools
Email: muhammadyasir85a@gmail.com
Location: Mianwali, Pakistan
University: Namal University Mianwali