Integrating physiological signals for enhanced sleep apnea diagnosis with SleepNet Prashant Hemrajani, Vijaypal Singh Dhaka, Geeta Rani, Sahil Verma, Kavita, Marcin Woźniak, Jana Shafi, Muhammad Fazal Ijaz Scientific Reports, 2025 Sleep apnea, a prevalent respiratory disorder, poses significant health risks, including cardiovascular complications and behavioral issues, if left untreated. Traditional diagnostic methods like polysomnography, although effective, are often expensive and inconvenient. SleepNet addresses these issues by introducing a new multimodal approach tailored for precise sleep apnea detection. At its core, the framework utilizes a fusion of one-dimensional convolutional neural networks (1D-CNN) and bidirectional gated recurrent units (Bi-GRU) to analyze single-lead electrocardiogram (ECG) recordings, yielding an accuracy of 95.08%. When the model is enriched with additional physiological signals-namely nasal airflow and abdominal respiratory effort-the performance further rises modestly to 95.19%. This multimodal strategy surpasses the performance of existing unimodal approaches, yielding enhanced sensitivity and specificity rates of 96.12% and 93.45%, respectively. When compared to previous studies, SleepNet represents a substantial leap forward in diagnostic efficacy, showcasing the transformative potential of integrating multiple data streams for sleep apnea detection. The results highlight the promise of deep learning methodologies in advancing this domain and lay a robust foundation for subsequent research.
NoRef-CLIP: Image Quality Assessment via Prompted Vision-Language Models Nandita Rawat, Geeta Rani 2025 International Conference on Emerging Trends in Networks and Computer Communications Etncc 2025 Proceedings, 2025 In many applications, Image Quality Assessment (IQA) is essential, particularly when accessing high-quality reference images is limited or challenging. Novel approaches to this task have been made possible by recent developments in self-supervised learning and vision-language models such as CLIP. The new self-supervised NR-IQA technique NoRefCLIP, which uses CLIP for image-text alignment, is examined in this paper. The methodology, architecture, experimental findings, and upcoming issues are all covered in detail in the report. Predicting the perceptual quality of an image without having access to a perfect, distortion-free reference image is known as No-Reference Image Quality Assessment (NR-IQA), and it is an essential task in computer vision. Unlike Full-Reference (FRIQA) and Reduced-Reference (RR-IQA) methods, which compare a degraded image to an original or partially available version, NR-IQA operates in a blind setting-making it significantly more challenging and more practical for real-world applications. By estimating how a normal observer would evaluate the visual quality of an image impacted by noise, blur, compression artifacts, or other distortions, NR-IQA aims to replicate human visual perception. This is especially crucial in situations where reference images may be hard to find or impractical to obtain, such as image enhancement, compression optimization, content delivery, and quality monitoring in fields like medical imaging, satellite imaging, and video streaming. Opinion-aware models, which are trained on datasets labeled with human Mean Opinion Scores (MOS) or other subjective quality annotations, comprise the majority of current NR-IQA techniques. While these models can perform well within the domains they are trained on, they suffer from key limitations: excessive reliance on subjective and expensive annotated data, Scalability problems when applied to new tasks or datasets without ground truth, as well as limited generalization to hidden distortion types or domains. Opinion-unaware, self-supervised, or unsupervised approaches that seek to estimate perceptual quality without depending on human-provided scores have become more popular as a result of these difficulties. These methods create new opportunities for flexible and scalable IQA systems that can operate in a variety of settings with limited annotations. To address this we propose NoRefCLIP (Quality-aware CLIP), an unsupervised, opinion-blind, CLIP-based system, which does not rely on any human-supplied quality annotations. In order to allow CLIP to generate image representations that are sensitive to quality degradations, NoRefCLIP presents a quality-aware image-text alignment mechanism. Synthetic degradations of varying degrees of severity are applied to pristine images first. After that, the model is trained to rank these deteriorated images according to how closely they resemble quality-relevant antonym text prompts. Consistent representations for images with comparable content and degradation levels are also encouraged.
Tomato TransDeepLab: A Robust Framework for Tomato Leaf Segmentation, Disease Severity Prediction, and Crop Loss Estimation Ankita Gangwar, Geeta Rani, Vijaypal Singh Dhaka IEEE Access, 2025 This research aims to reduce crop loss in tomato crops by developing an effective disease detection and segmentation system. To meet this objective, the researchers prepared three annotated real-world datasets consisting of tomato leaf images, including samples with simple and complex backgrounds, and young plants aged two to four weeks. A transformer-based deep learning model, ‘Tomato TransDeepLab’, is proposed for accurately segmenting multiple and minute disease lesions on a single leaf. The model demonstrates high segmentation precision even with complex backgrounds and indistinct lesion boundaries. It achieves a maximum Intersection over Union (IoU) of 89.09% and accuracy of 98.17%. It also recorded the lowest training time of 989 minutes. Comparative analysis with state-of-the-art architectures including U-Net, ResUNet, DeepLabV3, and Tuned DeepLabV3 + across all three datasets confirms its superior performance. Furthermore, the model is employed to assess disease severity and estimate crop loss using a severity scale designed by plant pathologists. This scale links lesion area to severity level and crop loss. Results indicate a crop loss between 1 to 5% for Datasets 1 and 2. The proposed approach offers an automated solution for disease segmentation, severity assessment, and crop loss estimation in tomato crops.
Adverse Drug Event Detection using NLP Kasvi Soni, Geeta Rani, Vijaypal Singh Dhaka, Sushma Hans 2025 International Conference on Emerging Trends in Networks and Computer Communications Etncc 2025 Proceedings, 2025 Adverse Drug Events are a major cause of morbidity and mortality globally. It leads to more than <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1. 3}$</tex> million emergency hospitalizations per year in the United States alone. Timely identification of Adverse Drug Event from text sources (e.g. social media, clinical reports) is essential to avoid significant health consequences and enhance patient outcomes. Traditional manual monitoring procedures for ADE detection are labourintensive and are more prone to error. Machine learning algorithms help to overcome these obstacles by facilitating automated, scalable analysis of vast textual data sources with higher accuracy. While prior research has explored machine learning for Adverse Drug Event detection, the optimal combination of feature extraction techniques and classification models remains understudied, leading to non-optimal outcomes in real-world applications. Previous works have focused on single methods, like Bag-of-Words with Support Vector Machines or TF-IDF with Logistic Regression, but comparative evaluation of these methods with hybrid feature extraction methods is less explored. Moreover, computational efficiency is rarely considered. This research performs a detailed comparison of various feature extraction techniques including Bag of Words, TF-IDF, Word2Vec, and hybrid techniques with Logistic Regression and Random Forest algorithms. The performance was evaluated in terms of accuracy, precision, recall, F1-score, training, and testing time. The findings showed that Bag of Words with Random Forest provides the best accuracy of 89.88 %. Here, hybrid approaches such as an integration of TF-IDF with Word2Vec were not superior to single methods indicating that more straightforward strategies are better suited for Adverse Drug Detection tasks.
LSEG: Lung Segmentation for Pulmonary Disease Affected Chest Radiographs Ankit Misra, Geeta Rani, Vijaypal Singh Dhaka Proceedings 2024 Joint International Conference on Digital Arts Media and Technology with Ecti Northern Section Conference on Electrical Electronics Computer and Telecommunications Engineering Ecti Damt and Ncon 2024, 2024
Exoskeleton Pysiotherapy and Assistive Robotic Arm Pradeep Surya Dadi, Geetha Rani K, Sathish Kumar P. J, Rajasekar B, Surendran R 2nd International Conference on Sustainable Computing and Smart Systems Icscss 2024 Proceedings, 2024
A Survey: Impact of Brain Stress on EEG Signals 12th Indiacom 5th International Conference on Computing for Sustainable Global Development Indiacom 2018, 2018
AI Based System for Mass Screening of Kidney Diseases Using Ultrasound Images G Rani, N Kundu, VS Dhaka, A Sharda 2026 5th Asia Conference on Algorithms, Computing and Machine Learning … , 2026 2026
Machine learning approach to gait analysis for Parkinson’s disease detection and severity classification R Mittal, N Agarwal, M Dubey, V Pathak, P Shukla, G Rani, E Vocaturo, ... Frontiers in Robotics and AI 12, 1623529 , 2025 2025 Citations: 1
Towards an Integrated AI Pipeline for Disease Diagnosis and Quality Assessment in Coffee Production G Rani, VS Dhaka, T Ruga, E Vocaturo, E Zumpano 2025 IEEE International Conference on Big Data (BigData), 7596-7605 , 2025 2025 Citations: 1
Innovations in intelligent automation and evolving human-Robot Interaction G Rani International Conference on Computer Science and Communication Engineering … , 2025 2025
3D-Printed AI-Powered Robot With Secure Data Handling and Return to Home Capabilities R Mittal, D Khandelwal, KH Nemade, G Rani, VS Dhaka, V Pathak, ... IEEE Access 13, 189915-189925 , 2025 2025
Tomato TransDeepLab: A Robust Framework for Tomato Leaf Segmentation, Disease Severity Prediction, and Crop Loss Estimation A Gangwar, G Rani, VS Dhaka IEEE Access , 2025 2025 Citations: 5
Integrating physiological signals for enhanced sleep apnea diagnosis with SleepNet P Hemrajani, VS Dhaka, G Rani, S Verma, Kavita, M Woźniak, J Shafi, ... Scientific Reports 15 (1), 31715 , 2025 2025 Citations: 8
NoRef-CLIP: Image Quality Assessment via Prompted Vision-Language Models N Rawat, G Rani 2025 International Conference on Emerging Trends in Networks and Computer … , 2025 2025
Adverse Drug Event Detection using NLP K Soni, G Rani, VS Dhaka, S Hans 2025 International Conference on Emerging Trends in Networks and Computer … , 2025 2025
Alzheimer Diagnosis and Cost Estimation Bot S Kumar, N Kundu, G Rani International Conference on Innovations in Computational Intelligence and … , 2025 2025
SpyBot: A Lightweight Robot that Works on LiDAR SLAM V Patel, G Rani International Conference on Innovations in Computational Intelligence and … , 2025 2025
Retrieval-Augmented Generation Approach for Media Search S Poply, A Chawla, V Vasudeva, G Rani International Conference on Innovations in Computational Intelligence and … , 2025 2025
Optimizing Image Reconstruction Techniques with Artificial Bee Colony (ABC) Algorithm U Pandey, M Kumari, G Rani, A Khare, P Srivastava, M Bhardwaj International Conference on Innovations in Computational Intelligence and … , 2025 2025
Neural Genome Sequence Encoding for Downy Mildew Pathogen Detection: A Spatial Pyramid and Dual-Attention Approach G Rani, A Misra, N Kundu, VS Dhaka, P Lather IEEE Access , 2025 2025 Citations: 3
Lightweight and hybrid transformer-based solution for quick and reliable deepfake detection G Rani, A Kothekar, SG Philip, VS Dhaka, E Zumpano, E Vocaturo Frontiers in big Data 8, 1521653 , 2025 2025 Citations: 6
Innovations in Computational Intelligence and Computer Vision: Proceedings of ICICV 2024, Volume 1 N Dey, T Perumal, JMRS Tavares, S Roy, D Sinwar Springer Verlag, Singapore , 2025 2025
Exploring the Impact of Convolutional Neural Networks on Brain Tumor Detection Accuracy Sonam, A Kumar, G Rani, Archit International Conference on Emerging Technologies: Micro to Nano, 883-895 , 2024 2024
Impact of Architectural Changes and Hyperparameters Finetuning on the Performance of Convolutional Neural Networks: Case Study Using Covid-19 Dataset U Pandey, N Kundu, G Rani, M Kumari International Conference on Emerging Technologies: Micro to Nano, 907-925 , 2024 2024
Selecting a Base Deep Learning Model for Tomato Leaf Disease Classification A Gangwar, G Rani, VS Dhaka, S Khandelwal 2024 International Conference on Emerging Trends in Networks and Computer … , 2024 2024 Citations: 1
A deep learning based system for automatic sorting and quality grading of citrus fruits N Kundu, G Rani, VS Dhaka 2024 International Conference on Emerging Trends in Networks and Computer … , 2024 2024 Citations: 3
MOST CITED SCHOLAR PUBLICATIONS
A survey of deep convolutional neural networks applied for prediction of plant leaf diseases VS Dhaka, SV Meena, G Rani, D Sinwar, MF Ijaz, M Woźniak Sensors 21 (14), 4749 , 2021 2021.0 Citations: 493
IoT and interpretable machine learning based framework for disease prediction in pearl millet N Kundu, G Rani, VS Dhaka, K Gupta, SC Nayak, S Verma, MF Ijaz, ... Sensors 21 (16), 5386 , 2021 2021.0 Citations: 233
Disease detection, severity prediction, and crop loss estimation in MaizeCrop using deep learning N Kundu, G Rani, VS Dhaka, K Gupta, SC Nayaka, E Vocaturo, ... Artificial intelligence in agriculture 6, 276-291 , 2022 2022.0 Citations: 150
Deep learning for enhanced brain tumor detection and classification M Agarwal, G Rani, A Kumar, P Kumar, R Manikandan, AH Gandomi Results in Engineering 22, 102117 , 2024 2024.0 Citations: 144
AI-based yield prediction and smart irrigation D Sinwar, VS Dhaka, MK Sharma, G Rani Internet of Things and Analytics for Agriculture, Volume 2, 155-180 , 2019 2019.0 Citations: 128
Diabetes prediction using artificial neural network N Pradhan, G Rani, VS Dhaka, RC Poonia Deep learning techniques for biomedical and health informatics, 327-339 , 2020 2020.0 Citations: 124
Role of internet of things and deep learning techniques in plant disease detection and classification: A focused review VS Dhaka, N Kundu, G Rani, E Zumpano, E Vocaturo Sensors 23 (18), 7877 , 2023 2023.0 Citations: 76
Seeds classification and quality testing using deep learning and YOLO v5 N Kundu, G Rani, VS Dhaka Proceedings of the international conference on data science, machine … , 2021 2021.0 Citations: 49
Applying deep learning-based multi-modal for detection of coronavirus G Rani, MG Oza, VS Dhaka, N Pradhan, S Verma, JJPC Rodrigues Multimedia Systems 28 (4), 1251-1262 , 2022 2022.0 Citations: 48
Efficient deep learning based hybrid model to detect obstructive sleep apnea P Hemrajani, VS Dhaka, G Rani, P Shukla, DP Bavirisetti Sensors 23 (10), 4692 , 2023 2023.0 Citations: 47
Comparative performance assessment of deep learning based image steganography techniques V Himthani, VS Dhaka, M Kaur, G Rani, M Oza, HN Lee Scientific Reports 12 (1), 16895 , 2022 2022.0 Citations: 46
A deep reinforcement learning technique for bug detection in video games G Rani, U Pandey, AA Wagde, VS Dhaka International Journal of Information Technology 15 (1), 355-367 , 2023 2023.0 Citations: 44
Spatial feature and resolution maximization GAN for bone suppression in chest radiographs G Rani, A Misra, VS Dhaka, E Zumpano, E Vocaturo Computer Methods and Programs in Biomedicine 224, 107024 , 2022 2022.0 Citations: 41
A comparative analysis of deep learning models applied for disease classification in bell pepper N Kundu, G Rani, VS Dhaka 2020 sixth international conference on parallel, distributed and grid … , 2020 2020.0 Citations: 40
Role of artificial neural networks in predicting design and efficiency of dye sensitized solar cells N Tomar, G Rani, VS Dhaka, PK Surolia International Journal of Energy Research 46 (9), 11556-11573 , 2022 2022.0 Citations: 37
Optimized contrast enhancement for tumor detection VS Monika Agarwal, Geeta Rani International Journal of Imaging Systems and Technology , 0 Citations: 37
Time and space efficient multi-model convolution vision transformer for tomato disease detection from leaf images with varied backgrounds A Gangwar, V Dhaka, G Rani, S Khandelwal, E Zumpano, E Vocaturo Computers, Materials, & Continua 79 (1), 117 , 2024 2024.0 Citations: 36
KUB-UNet: segmentation of organs of urinary system from a KUB X-ray image G Rani, P Thakkar, A Verma, V Mehta, R Chavan, VS Dhaka, RK Sharma, ... Computer Methods and Programs in Biomedicine 224, 107031 , 2022 2022.0 Citations: 36
A multi-modal bone suppression, lung segmentation, and classification approach for accurate COVID-19 detection using chest radiographs G Rani, A Misra, VS Dhaka, D Buddhi, RK Sharma, E Zumpano, ... Intelligent Systems with Applications 16, 200148 , 2022 2022.0 Citations: 35
Transforming view of medical images using deep learning N Pradhan, VS Dhaka, G Rani, H Chaudhary Neural Computing and Applications 32 (18), 15043-15054 , 2020 2020.0 Citations: 34
