Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Maria Tamoor has a total of 40 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Automated cobb’s angle measurement for scoliosis diagnosis using deep learning techniques

    by Asbah Khalid, Asma Ahmad Farhan, Kashif Zafar, Maria Tamoor

    Purpose: Scoliosis is a widespread musculoskeletal disorder of bending and twisting of spine. In this, spine curves to the side and in severe cases it can twist and take several bends. For the diagnosis and treatment of scoliotic patients, the Cobb’s angle is a critical marker of the body’s curvature. Early detection and proper treatment of scoliosis help individuals maintain a higher quality of life. There are many researches that have been conducted to automate the manual measurement of the angle and every investigation has their own limitations. Approach: This paper presents a method for precisely measuring the Cobb’s angle using deep learning based techniques. Mainly it comprises of feature enhancement of augmented dataset, a bespoke code for landmark estimation on the spine, segmentation model based on the U-Net architecture and a custom code for Cobb’s angle measurement. These measured angles are compared to the given angles for the segmentation on X-ray images. Results: The proposed technique offers an automated way for precisely measuring this angle with an overall accuracy of 97 % and Root Mean Square Error of 4.99 , which can lower the variability, facilitate early detection, accurate diagnosis, and monitoring of scoliosis progression. Additionally, it aids in the treatment planning and evaluating treatment outcomes. Conclusion: The combination of deep learning techniques, accurate landmark estimation, and segmentation has enabled to develop an automated system that can consistently and objectively measure the Cobb’s angle with high accuracy. Healthcare professionals can enhance the quality of care and can improve long-term outcomes for individuals with scoliosis using this method.


  • Open Access

    DiscHAR: A Discrete Approach to Enhance Human Activity Recognition in Cyber Physical Systems: Smart Homes

    by Ishrat Fatima, Asma Ahmad Farhan, Maria Tamoor, Shafiq Ur Rehman, Hisham Abdulrahman Alhulayyil, Fawaz Tariq

    The main challenges in smart home systems and cyber-physical systems come from not having enough data and unclear interpretation; thus, there is still a lot to be done in this field. In this work, we propose a practical approach called Discrete Human Activity Recognition (DiscHAR) based on prior research to enhance Human Activity Recognition (HAR). Our goal is to generate diverse data to build better models for activity classification. To tackle overfitting, which often occurs with small datasets, we generate data and convert them into discrete forms, improving classification accuracy. Our methodology includes advanced techniques like the R-Frame method for sampling and the Mixed-up approach for data generation. We apply K-means vector quantization to categorize the data, and through the elbow method, we determine the optimal number of clusters. The discrete sequences are converted into one-hot encoded vectors and fed into a CNN model to ensure precise recognition of human activities. Evaluations on the OPP79, PAMAP2, and WISDM datasets show that our approach outperforms existing models, achieving 89% accuracy for OPP79, 93.24% for PAMAP2, and 100% for WISDM. These results demonstrate the model’s effectiveness in identifying complex activities captured by wearable devices. Our work combines theory and practice to address ongoing challenges in this field, aiming to improve the reliability and performance of activity recognition systems in dynamic environments.


  • Open Access

    Investigating the Taxonomy of Character Recognition Systems: A Systematic Literature Review

    by Asma Naseer, Maria Tamoor, Nasser Allheeib, Summrina Kanwal

    Taxonomy, a scientific and systematic categorization of elements, has been extensively applied in various domains, including data grids, data mining tasks, and network systems. However, until now, there has been a notable absence of research exploring the taxonomy of Character Recognition (CR) Systems. CR, the process of identifying characters in image format and associating them with their respective ASCII or Unicode, presents varied mechanisms for different phases of the recognition process. Our study centers around the development of a taxonomy for CR, exploring both contemporary trends and obstacles within the domain. We pivot to CR , systematically examining pivotal technologies, application scenarios leveraging state-of-the-art machine learning models, and associated services within diverse contexts. The narrative encompasses an exploration of the challenges and constraints inherent in CR systems. Through a systematic literature review, we scrutinize the fundamental technologies, practical applications, and research trajectories in CR, pinpointing burgeoning developments and avenues for further investigation. This diversity enables us to classify character recognition systems into groups and subgroups, paving the way for an intensive taxonomy of these systems by delveing into existing character recognition systems, identifying similarities and differences. Ultimately, we propose a taxonomy of character recognition systems, offering a novel perspective on this domain. A rigorous selection process was undertaken, identifying and categorizing 96 papers published between 2018 and 2024 according to predefined criteria. The findings are meticulously analyzed, offering a panoramic taxonomy of character recognition system in various contexts.


  • Open Access

    A Novel Technique for Optimizing Volumetric Avatars: From 2D Images to Lightweight 3D Models

    by Kinza Shahzad, Asma Naseer, Aamir Wali, Maria Tamoor

    In the world of digital avatars, three-dimensional (3D) human models are utilized to replicate real-world appearance and movements with a higher degree of realism compared to traditional 2D representations. Currently, extensive person-specific data capturing and 3D artists are required to create photorealistic avatars of existing people. Determining the 3D pose and articulation of an avatar from 2D images requires complex algorithms, a capability commonly found in the Visual Effects (VFX) industry but not readily available elsewhere. Hence, this study proposes an approach to generate 3D avatar from a 2D image of a person with uncanny resemblance and an accurate depiction of the subject’s likeness. The proposed approach, PIFu+CycleGAN, combines Pixel-aligned Implicit Function (PIFu) and cycleGAN for textured avatar construction. PIFu is specifically used to capture intricate details and arbitrary topology. Additionally, the hourglass architecture-based module is utilized for T-pose estimation for predicting the initial geometry and shape. Evaluating the proposed approach on the benchmark RenderPeople dataset, it outperforms the state-of-the-art models with values of 1.53 and 1.50 for the Chamfer and P2S distance, respectively. This indicates the creation of high-quality 3D meshes with promising textures, which are of exceptional quality and suitable for animation.

  • Open Access

    Improving Mental Health with Machine Learning: Classification of Mental Disorders Using Local Data

    by Maria Tamoor, Haziq Ahmed, Ahmad Saad, Hamza Asaad, Muhammad Sulaiman, Mubashar Mushtaq

    Researchers are dedicated to providing effective solutions for serious problems such as mental health. This field of study ranges from understanding the impact of mental health ailments on individuals to identifying trends in mental health diseases based on various data points. Unfortunately, the journey from diagnosis to receiving a proper mental health regimen is costly and inaccessible to many. This necessitates the introduction of an affordable alternative that enables individuals with internet access to connect to a platform to determine whether they require the services of a mental health professional. Additionally, such a platform can assist mental health professionals in quickly identifying patients with mental health conditions. Our proposed solution leverages hidden trends in data obtained from 700 mental health patients at The local mental hospital in Pakistan. We have used this data to train a machine learning classifier capable of distinguishing between Psychotic and Neurotic disorders. Psychotic disorders are a group of mental health conditions characterized by a loss of touch with reality whereas neurotic disorders are a group of mental health conditions characterized by excessive worry, anxiety, or stress that does not involve a loss of touch with reality. Our model achieved an overall accuracy score of 73%, indicating successful prediction of the correct diagnosis in 73 out of 100 pipeline tests. While these results may not seem significant, an increase in data points would surely allow for further improvements in these results and the overall solution we propose.

  • Open Access

    Automated Cobb’s Angle Measurement for Scoliosis Diagnosis Using Deep Learning Technique

    by Asbah Khalid, Asma Ahmad Farhan, Kashif Zafar, Maria Tamoor

    BACKGROUND: Scoliosis is a widespread musculoskeletal disorder of bending and twisting of spine. In this medical ailment, spine curves to the side and even in severe cases it can twist and can take several bends. For the diagnosis and treatment of scoliotic patients, the Cobb’s angle is a critical marker of the body’s curvature. OBJECTIVE: There are many researches that have been conducted to automate the manual measurement of the angle and every investigation has their own limita- tions. METHODS: This paper presents a method for precisely measuring the Cobb’s an- gle using deep learning based techniques. Mainly it comprises of feature enhance- ment of augmented dataset, a bespoke code for landmark estimation on the spine, segmentation model based on the U-Net architecture and a custom code for Cobb’s angle measurement. These measured angles are then compared to the given angles for the segmentation on biomedical (X-ray) images. RESULTS: The findings demonstrate that the proposed technique offers an auto- mated and impartial way for precisely measuring this angle with an overall accu- racy of 97% and Root Mean Square Error (RMSE) of 4 . 99, which can lower the variability, facilitate early detection, accurate diagnosis, and monitoring of scolio- sis progression. Additionally, it aids in the treatment planning and evaluating treat- ment outcomes. CONTRIBUTION: By leveraging the applications of Cobb’s angle measurement, healthcare professionals can enhance the quality of care and can improve long-term outcomes for individuals with scoliosis.

  • Open Access

    Smart Agriculture: IoT and Deep Learning for Precision Crop Management

    by Hooriya Najeeb, Asma Naseer, Maria Tamoor

    Pakistan is an agricultural country, and an exporter of crops in many countries. The crop production is an important factor to earn revenue for the country. But the agriculture of Pakistan is still not revolutionized and no modern system has been deployed commonly. Whereas the farmers face di ffi culty in growing some crops because those crops needs more care and calculated steps. Moreover the farmers face di ffi culty at di ff erent stages of the crop growth like to check fertility of soil and if the soil is non-fertile then which chemical property is lacking. The objective of this research is to provide digitize the agriculture of Pakistan. An emerg ing technology IoT and Deep learning based system will be developed. The IoT part contains some devices like sensors,gateways and communication technologies whereas the software part consists of Deep learning models that can make predictions about the fertility of the soil. The motive of this system is to provide best results in term of accuracy. Whereas the dataset used is publicly available on kaggle but it was of non-fertile values, so data augmentation is also performed to generate data for fertile values. We have implemented LSTM,RNN,CNN among them LSTM performed best for the prediction of soil fertility.

  • Open Access

    A Novel Hashcode-based Duplication Reduction via Thresholding Approach for Large-scale Web Documents

    by Sana Ejaz, Asma Naseer, Asma Ahmad, Maria Tamoor, Samina Naz

    Modern search engines encounter a significant challenge when it comes to handling duplicate and nearly identical web pages, particularly during the indexing process for vast amounts of web content. This issue can lead to slow search results and increased costs due to the accumulation of storage space necessary for storing indexes. To tackle this issue, different techniques have been proposed to find similar websites. However, it has long been a challenge in research to distinguish between web pages. In the current study, sentence-level features i.e., hashcode and thresholding are used to determine the nearly identical web pages. We employ an adaptive threshold that enables the application of our model in both large- and small-scale settings. The benchmark datasets consisting of Shakespeare’s collections, free text, job descriptions, and Reuters-21578 are used to test the proposed approach. With an accuracy of 0.99 and an F1-score of 0.97, the proposed technique outperforms existing methods.

  • Open Access

    Advancements in Query-Based Tabular Data Retrieval: Detecting Image Data Tables and Extracting Text using Convolutional Neural Networks

    by Hina Tufail, Asma Naseer, Maria Tamoor, Abbas Raza Ali

    The detection and recognition of tables in image-based documents is a challenging and essential task for automated data extraction and anal- ysis. This paper introduces a methodology that utilizes Convolutional Neural Networks (CNNs) to address this problem. By harnessing the ro- bust visual recognition abilities of CNNs, the proposed method effectively detects and extracts tables from image-based documents. Additionally, an Encoder-Dual Decoder (EDD) architecture is employed to extract the textual content from table cells, enabling the transformation of visual data into structured, machine-readable formats. This structured data is then further processed using a Vector Space Model (VSM)-based language modeling technique for query-based table retrieval. The methodology’s accuracy is assessed by analyzing the spatial relationships between the extracted cells and comparing them against a pre-defined table structure. The results provide an evidence to the effectiveness of the proposed ap- proach, with an impressive 85% accuracy rate and a corresponding error rate of 15%. This accurate extraction and recognition of tables from image-based documents can significantly enhance applications in data analysis, information retrieval, and knowledge extraction.


  • Open Access

    Two-Phase Segmentation Approach for Accurate Left Ventricle Segmentation in Cardiac MRI using Machine Learning

    by Maria Tamoor, Abbas Raza Ali, Philemon Philip, Ruqqayia Adil, Rabia Shahid, Asma Naseer

    Accurate segmentation of the Left Ventricle (LV) holds substantial importance due to its implications in disease detection, regional analysis, and the development of complex models for cardiac surgical planning. CMR is a golden standard for diagnosis of serveral cardiac diseases. LV in CMR comprises of three distinct sections: Basal, Mid-Ventricle, and Apical. This research focuses on the precise segmentation of the LV from Cardiac MRI (CMR) scans, joining with the capabilities of Machine Learning (ML). The central challenge in this research revolves around the absence of a set of parameters applicable to all three types of LV slices. Parameters optimized for basal slices often fall short when applied to mid-ventricular and apical slices, and vice versa. To handle this issue, a new method is proposed to enhance LV segmentation. The proposed method involves using distinct sets of parameters for each type of slice, resulting in a two-phase segmentation approach. The initial phase categorizes images into three groups based on the type of LV slice, while the second phase aims to segment CMR images using parameters derived from the preceding phase. A publicly available dataset (Automated Cardiac Diagnosis Challenge (ACDC)) is used. 10-Fold Cross Validation is used and it achieved a mean score of 0.9228. Comprehensive testing indicates that the best parameter set for a particular type of slice does not perform adequately for the other slice types. All results show that the proposed approach fills a critical void in parameter standardization through a two-phase segmentation model for the LV, aiming to not only improve the accuracy of cardiac image analysis but also contribute advancements to the field of LV segmentation.