- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Radiomics and Machine Learning in Medical Imaging
- Remote-Sensing Image Classification
- Face recognition and analysis
- Face and Expression Recognition
- Medical Image Segmentation Techniques
- Maternal and fetal healthcare
- Topic Modeling
- Cardiac, Anesthesia and Surgical Outcomes
- Microscopic Colitis
- Inflammatory Bowel Disease
- Vaccine Coverage and Hesitancy
- Anesthesia and Sedative Agents
- AI in cancer detection
- Maternal and Perinatal Health Interventions
- Cardiac Health and Mental Health
- Satellite Image Processing and Photogrammetry
- Biomedical Text Mining and Ontologies
- Spam and Phishing Detection
- Handwritten Text Recognition Techniques
- COVID-19 epidemiological studies
- COVID-19 Clinical Research Studies
- Pregnancy and preeclampsia studies
- Viral Infections and Outbreaks Research
Ayub Medical College
2025
Government College University, Faisalabad
2018-2024
Sindh Institute of Urology and Transplantation
2023-2024
Military Hospital
2021-2023
National University of Medical Sciences
2018-2023
University of Helsinki
2023
Combined Military Hospital
2018-2022
Texas Tech University
2020-2021
The University of Texas of the Permian Basin
2021
Texas Tech University Health Sciences Center
2020-2021
Image classification has gained lot of attention due to its application in different computer vision tasks such as remote sensing, scene analysis, surveillance, object detection, and image retrieval. The primary goal is assign the class labels images according contents. applications analysis sensing are important they used various applied domains military civil fields. Earlier approaches for based on low-level feature representations color- texture-based features. Vector Locally Aggregated...
The requirement for effective image search, which motivates the use of Content-Based Image Retrieval (CBIR) and search similar multimedia contents on basis user query, remains an open research problem computer vision applications. application domains Bag Visual Words (BoVW) based representations are object recognition, classification content-based analysis. Interest point detectors quantized in feature space final histogram or signature do not retain any detail about co-occurrences features...
Automated recognition of handwritten characters and digits is a challenging task. Although significant amount literature exists for automatic English other major languages in the world, there wide research gap due to lack Urdu language. The variations writing style, shape size individual similarities with add complexity accurate classification characters. Deep neural networks have emerged as powerful technology automated character patters object images. deep are known provide remarkable...
The recent development in the technology has increased complexity of image contents and demand for classification becomes more imperative. Digital images play a vital role many applied domains such as remote sensing, scene analysis, medical care, textile industry crime investigation. Feature extraction representation is considered an important step analysis it affects performance. Automatic open research problem pattern recognition applications. Bag-of-Features (BoF) model commonly used to...
Recently, face datasets containing celebrities photos with facial makeup are growing at exponential rates, making their recognition very challenging. Existing methods rely on feature extraction and reference reranking to improve the performance. However images carry inherent ambiguity due artificial colors, shading, contouring, varying skin tones, task more difficult. The problem becomes confound as alters bilateral size symmetry of certain components such eyes lips affecting distinctiveness...
Agile software development has large success rate due to its benefits and promising nature but natively where the size of project is small. Requirement engineering (RE) crucial as in each life cycle, “Requirements” play a vital role. Though agile provides values customer’s business needs, changing requirement, interaction, we also have face impediments agile, many which are related requirement challenges. This article aims find out challenges being faced during projects. Many research...
Due to the influx of advancements in technology and increased simplicity communication through emails, there has been a severe threat global economy security due upsurge volume unsolicited During training models, high-dimensional redundant datasets may reduce classification results model high memory costs computation. An important data processing technique is feature selection which helps selecting relevant features subsets information from dataset. Therefore, choosing efficient techniques...
The Bag-of-Visual-Words (BoVW) model is widely used for image classification, object recognition and retrieval problems. In BoVW model, the local features are quantized 2-D space represented in form of order-less histogram visual words. classification performance suffers due to representation image. This paper presents a novel that incorporates spatial information inverted index model. added by calculating global relative orientation words rotation invariant manner. For this, we computed...
Respiratory viruses known as coronaviruses infect people and cause death. The multiple crown-like spikes on the virus’s surface give them name “corona”. pandemic has resulted in a global health crisis it is expected that every year we will have to fight against different COVID-19 variants. In this critical situation, existence of vaccinations provides hope for mankind. Despite severe vaccination campaigns recommendations from experts government, perceptions regarding risks share their views...
The amount of multimedia content is growing exponentially and a major portion uses images video. Researchers in the computer vision community are exploring possible directions to enhance system accuracy reliability, these main requirements for robot vision-based systems. Due change facial expressions wearing masks or sunglasses, many face recognition systems fail recognizing decreases scenarios. In this work, we contribute real time surveillance framework using Raspberry Pi CNN...
Aging affects left and right half face differently owing to numerous factors such as sleeping habits, exposure sun light, weaker muscles of one side face. In computer vision, age a given image is estimated using features that are correlated with age, moles, scars, wrinkles. this study we report the asymmetric aging sides images its impact on accurate estimation. Left symmetric faces were perceived younger while older when presented state‐of‐the‐art estimator. These findings show facial an...
As digital images play a vital role in multimedia content, the automatic classification of is an open research problem. The Bag Visual Words (BoVW) model used for image classification, retrieval and object recognition problems. In BoVW model, histogram visual words computed without considering spatial layout 2-D space. performance suffers due to lack information about details image. Spatial Pyramid Matching (SPM) popular technique that computes However, SPM not rotation-invariant does allow...
The classification of high-resolution satellite images is an open research problem for computer vision community. In last few decades, the Bag Visual Word (BoVW) model has been used images. BoVW model, orderless histogram visual words without any spatial information as image signature. performance suffers due to this nature and addition clues are reported beneficial scene geographical Most representations that can compute not invariant rotations. A rotation representation considered one main...
Background: Coronary artery disease (CAD) is a leading cause of global morbidity and mortality, resulting in impaired blood flow to the myocardium, myocardial infarction, heart failure. Major risk factors include hypertension, diabetes mellitus, hyperlipidemia, smoking, physical inactivity, advancing age. Despite burden CAD, limited data exists on its predictors low-resource settings, particularly South Asia. This study aimed assess frequency conventional for CAD among patients admitted...
Background: Diabetes mellitus (DM) is a leading non-communicable disease worldwide, with Pakistan ranking third globally in terms of diabetes prevalence. Chronic complications impose significant burden on healthcare systems, to increased morbidity, costs, and reduced quality life. Despite advancements treatment, substantial proportion patients develop that could be mitigated through early diagnosis, lifestyle modifications, adherence treatment. Identifying risk factors their association...
Postmenopausal women frequently experience urinary incontinence (UI), a condition often overlooked due to social stigma and lack of awareness. This study aimed assess the risk factors that contribute development in postmenopausal women. Objective: To determine associated with Methods: is cross-sectional conducted at Uro-Gynae Unit Ayub Teaching Hospital, Abbottabad, on 126 patients, following approval from Ethical Committee using non-probability consecutive sampling technique. Data was...
Background: Chronic kidney disease (CKD) is a progressive condition with rising global prevalence, often requiring long-term dialysis to sustain life. While extends survival, it imposes significant psychological and physical stress on both patients their caregivers. Mental health issues such as anxiety, depression, fatigue are common yet underrecognized consequences of this burden. Understanding the impact role social support essential improving quality life treatment outcomes for vulnerable...
Semantic taxonomies are powerful tools that provide structured knowledge to Natural Language Processing (NLP), Information Retreval (IR), and general Artificial Intelligence (AI) systems. These extensively used for solving rich problems such as textual entailment question answering. In this paper, we present a taxonomy induction system evaluate it using the benchmarks provided in Taxonomy Extraction Evaluation (TExEval2) Task. The task is identify hyponym-hypernym relations construct from...
Disruptive behavior is a prevalent threat to constructive online engagement. Covert behaviors, like trolling, are especially challenging detect automatically, because they utilize deceptive strategies manipulate conversation. We illustrate novel approach their detection: analyzing conversational structures instead of focusing only on messages in isolation. Building conversation analysis, we demonstrate that 1) actions and norms provide concepts for deeper understanding covert disruption, 2)...