Looking for a job

Zabeeh ullah Noor

@zabeehullahnoor

Message

AI and computer vision engineer at NUST who enhanced 3D Gaussian Splatting for novel view synthesis from sparse images.

Pakistan

Message

At NUST, I enhanced 3D Gaussian Splatting for few-shot novel view synthesis, reconstructing 3D scenes from 5–7 sparse images. I integrated DepthAnything-V2 and designed an improved loss function to support geometric consistency.

I also developed real-time vision systems, including tracking for a 6-DOF robotic arm and people and vehicle tracking with YOLO, ByteTrack, DeepSORT, and OpenCV. I developed VisionAgent, an image analysis platform combining object detection, scene understanding, and language-model reasoning.

At Independent AI Solutions, I developed computer vision systems for medical image segmentation, pneumonia detection, facial emotion recognition, OCR, and hand gesture recognition. Earlier, at MZ Developers, I built Android applications with backend services and REST APIs.

Experience

Work history, roles, and key accomplishments

Machine Learning & Computer Vision Developer

Independent AI Solutions

Feb 2022 - Mar 2024 (2 years 1 month)

Developed medical image segmentation models, pneumonia detection systems, facial emotion recognition, OCR systems, and real-time hand gesture recognition. Applied data preprocessing, augmentation, and model evaluation across multiple deep learning tasks.

Attention U Net EfficientNet ResNet CNN Bi LSTM CTC Mediapipe OpenCV YOLO scikit learn

Android & Backend Developer

MZ Developers

Mar 2020 - Jan 2022 (1 year 10 months)

Built full-stack Android applications with backend services and REST APIs using Django REST Framework. Developed real-time messaging and social feed systems, designed PostgreSQL schemas, and implemented CI/CD pipelines.

Android Django REST Framework PostgreSQL REST APIs GitHub Actions CI CD

Deep Learning & Computer Vision Engineer

NUST

Apr 2024 - Present (2 years 5 months)

Enhanced 3D Gaussian Splatting for few-shot novel view synthesis and developed real-time vision-based tracking systems integrated with a 6-DOF robotic arm. Implemented people tracking, custom YOLO detection, and multi-lane vehicle counting systems.