Skip to content
View Hamza-Maqsood1's full-sized avatar

Block or report Hamza-Maqsood1

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Hamza-Maqsood1/README.md

Hi, I'm Hamza Maqsood πŸ‘‹

πŸŽ“ BS Artificial Intelligence - University of Management & Technology, Lahore (2022–2026) πŸ”¬ ML Engineer | Computer Vision Researcher | Robotics Enthusiast πŸ“„ Published: "Vision-only deep learning baseline for Meme-based humor classification" NBC 2026, Namal University (Track: Technology & Innovation Β· Abstract ID: T4.02) πŸ“ Lahore, Pakistan | Applying for MS in AI


🧠 Research Interests

  • Medical Image Analysis & Clinical AI (Diabetic Retinopathy, Healthcare Systems)
  • Multimodal & Vision-only Learning (CLIP, ResNet, Transfer Learning)
  • Explainable AI in Healthcare (SHAP, Interpretability)
  • Autonomous Systems & Robotics (ROS, Embedded AI)

πŸ“„ Publication

Vision-only deep learning baseline for Meme-based humor classification Hamza Maqsood - University of Management and Technology, Lahore πŸ“˜ Published in: Namal Business Conference 2026 Abstract Book (NBC 2026) 🏷️ Track: Technology and Innovation | Abstract ID: T4.02 πŸ”‘ Keywords: Visual Humor Classification Β· Meme Analysis Β· CNN Β· ResNet18 Β· Image-Based Learning πŸ”— View in Abstract Book

A deep learning framework for visual humor classification using fine-tuned ResNet18. Vision-only approach using ImageNet transfer learning addresses a gap where prior research focused on text or multimodal methods.


πŸš€ Featured Projects

Project Domain Highlight
Diabetic Retinopathy Detection Medical CV 95.95% val accuracy Β· MobileNetV2 Β· 7-class severity grading
CLIP Visual Semantic Similarity Multimodal AI 91.9% accuracy Β· Hard Negative Sampling Β· Focal Loss
IoT Healthcare Security Security + XAI SHAP Explainability Β· Ensemble Learning
Smart Email Agent NLP TF-IDF Β· Logistic Regression Β· Chainlit UI

πŸ€– Hardware & Robotics

  • Robotic Arm Prototype - designed and built with Arduino + servo control
  • Autonomous Systems - sensor integration with Raspberry Pi
  • ROS (Robot Operating System) - hands-on robotics middleware experience
  • Edge AI - ML inference deployment on embedded hardware

πŸ› οΈ Tech Stack

Languages: Python Β· C++ ML/DL: TensorFlow Β· PyTorch Β· Scikit-learn Β· OpenCV Architectures: ResNet18 Β· MobileNetV2 Β· CLIP Β· CNN NLP: NLTK Β· TF-IDF Β· Transformers Robotics: ROS Β· Arduino Β· Raspberry Pi Cloud: Google Cloud Platform (Certified) Β· Microsoft Azure AI Services Tools: Jupyter Β· Git Β· SHAP Β· Chainlit


πŸ“œ Certifications

  • πŸ† Google Cloud Certified - Professional Cloud Architect (Nov 2024)
  • ☁️ Develop NLP Solutions with Azure AI Services - Microsoft (Jan 2025)
  • πŸ€– Generative AI for Executives and Business Leaders - IBM (May 2024)
  • πŸ€– NLP Building Chatbots using AI - Udemy (Jan 2024)
  • πŸ“ How to write a research paper - Coursera (Jan 2025)
  • πŸ’» Research Paper Presentation - NBC(April 2026)
  • πŸ“ˆ Writing High-Quality Impact Factor Papers - NUM (April 2026)

πŸ“« Connect With Me

LinkedIn


πŸ’‘ Building at the intersection of AI, Healthcare, and Robotics. Published researcher. Open to MS programs, remote ML roles, and research collaborations.

Pinned Loading

  1. Meme-Humor-Classification-ResNet18 Meme-Humor-Classification-ResNet18 Public

    Vision-only meme humor classification using ResNet18. Published research - NBC 2026, Track: Technology & Innovation.

    Jupyter Notebook

  2. Adapting-CLIP-Embeddings-for-Visual-Semantic-Similarity Adapting-CLIP-Embeddings-for-Visual-Semantic-Similarity Public

    CLIP-based visual semantic similarity using hard negative sampling and focal loss. 91.9% validation accuracy.

    Jupyter Notebook

  3. Early-Detection-and-Classification-of-Diabetic-Retinopathy Early-Detection-and-Classification-of-Diabetic-Retinopathy Public

    CNN-based diabetic retinopathy detection with 95.95% accuracy. 7-class severity grading using MobileNetV2 and transfer learning.

    Jupyter Notebook

  4. scholar-ai scholar-ai Public

    Intelligent research agent - searches 250M+ academic papers, generates AI summaries with LLaMA 3.3 70B, exports PDF reports, and supports voice commands.

    HTML

  5. IoT_Healthcare_Security IoT_Healthcare_Security Public

    ML-based IoT healthcare security system with SHAP explainability, ensemble learning, and overfitting prevention via preprocessing.

    Jupyter Notebook

  6. Smart-Email-Agent Smart-Email-Agent Public

    NLP-based email classification agent using TF-IDF and Logistic Regression. Interactive UI built with Chainlit.

    Python