Research group, Department of Computer Engineering, CUI Lahore

Machine Vision and Intelligence

A camera sees pixels. We teach machines to find the person in them, track every joint, and learn from how they move: from gait analysis for remote patient care to facial expressions and limb signals.

See our work Meet the team

Mid-stance

Near hip 4°, near knee 5°

Gait cycle 30%

Fourteen keypoints through one stride, named by the standard phases of gait. The motion is an illustrative model, not patient data; the group's PhD research estimates it from video.

What we work on

Computer Vision

Finding people and objects, recognising them and outlining them: pose estimation for gait, facial micro-expressions, concealed objects in millimetre-wave images.

DetectionRecognitionSegmentation

Digital Image Processing

Enhancing images and extracting their features, from X-ray inspection of nuts to thermographic images of diabetic feet.

EnhancementFeature extraction

Applied Machine Learning

Learning from data across electrical and computer engineering: load forecasting, power outage risk, blood glucose prediction and EMG signals.

RegressionClassificationForecastingData visualization

The team

Full profiles and supervised theses

Dr. Ikram Ullah Khosa
Group lead. Associate Professor
Digital image processing, computer vision, machine learning. HEC approved PhD supervisor
Dr. Muhammad Babar Ali
Assistant Professor
Smart grid communication, IoT, low power wireless and sensor networks, propagation modelling, full-stack web development
Mr. Moazzam Ali Sahi
Lecturer
Digital image processing, computer architecture, software development, microcontroller-based system design
Ms. Wajeeha Khan
Lecturer, on leave
Embedded systems and computer architecture

33 papers since 2010

By members of the group, as listed on their CUI profiles.

Gait and movement

A hybrid neural network approach to gait analysis for remote patient care

ICT Express, 2026

A sequential-data approach for human gait parameter estimation and surgical event inference

IEEE Access, 2024

PhD in progress: human gait analysis and identification using pose estimation.

More recent papers

BWFER-YOLOv8: an enhanced cascaded framework for concealed object detectionApplied Sciences, 2025
Optimizing the impact of time domain segmentation techniques on upper limb EMG decoding using multimodal featuresPLOS One, 2025
Multi-stream deep convolution neural network with ensemble learning for facial micro-expression recognitionIEEE Access, 2023

All publications

Journal papers, 23Conference papers, 10Lines every 2 papers

Funded project

Facial landmark tracking in long and in-the-wild videos using deep learning

HEC, PKR 499,000, 2017 to 2019. Led by Dr. Ikram Ullah Khosa with Dr. Muhammad Haris Khan. Completed.

Students

Three PhD theses in progress, eleven MS theses since 2017, and over twenty-five vision and robotics projects with undergraduates, such as drowsy-driver detection and fruit quality testing.

Projects and theses

Contact

For research collaboration, MS and PhD supervision or final year projects, write to the group lead through the CUI message portal.

Message Dr. Ikram Ullah Khosa

Department of Computer Engineering
COMSATS University Islamabad, Lahore Campus
1.5 KM Defence Road, Off Raiwind Road, Lahore

+92 (42) 111-001-007