Nilanjan Dey

Md Farhan Ishmam, Md Sakib Hossain Shovon, M. F. Mridha et al.

The multimodal task of Visual Question Answering (VQA) encompassing elements of Computer Vision (CV) and Natural Language Processing (NLP), aims to generate answers to questions on any visual input. Over time, the scope of VQA has expanded from datasets focusing on an extensive collection of natural images to datasets featuring synthetic images, video, 3D environments, and various other visual inputs. The emergence of large pre-trained networks has shifted the early VQA approaches relying on feature extraction and fusion schemes to vision language pre-training (VLP) techniques. However, there is a lack of comprehensive surveys that encompass both traditional VQA architectures and contemporary VLP-based methods. Furthermore, the VLP challenges in the lens of VQA haven't been thoroughly explored, leaving room for potential open problems to emerge. Our work presents a survey in the domain of VQA that delves into the intricacies of VQA datasets and methods over the field's history, introduces a detailed taxonomy to categorize the facets of VQA, and highlights the recent trends, challenges, and scopes for improvement. We further generalize VQA to multimodal question answering, explore tasks related to VQA, and present a set of open problems for future investigation. The work aims to navigate both beginners and experts by shedding light on the potential avenues of research and expanding the boundaries of the field.

Md Abrar Jahin, M. F. Mridha, Zeyar Aung et al.

Geomagnetic storms, caused by solar wind energy transfer to Earth's magnetic field, can disrupt critical infrastructure like GPS, satellite communications, and power grids. The disturbance storm-time (Dst) index measures storm intensity. Despite advancements in empirical, physics-based, and machine-learning models using real-time solar wind data, accurately forecasting extreme geomagnetic events remains challenging due to noise and sensor failures. This research introduces TriQXNet, a novel hybrid classical-quantum neural network for Dst forecasting. Our model integrates classical and quantum computing, conformal prediction, and explainable AI (XAI) within a hybrid architecture. To ensure high-quality input data, we developed a comprehensive preprocessing pipeline that included feature selection, normalization, aggregation, and imputation. TriQXNet processes preprocessed solar wind data from NASA's ACE and NOAA's DSCOVR satellites, predicting the Dst index for the current hour and the next, providing vital advance notice to mitigate geomagnetic storm impacts. TriQXNet outperforms 13 state-of-the-art hybrid deep-learning models, achieving a root mean squared error of 9.27 nanoteslas (nT). Rigorous evaluation through 10-fold cross-validated paired t-tests confirmed its superior performance with 95% confidence. Conformal prediction techniques provide quantifiable uncertainty, which is essential for operational decisions, while XAI methods like ShapTime enhance interpretability. Comparative analysis shows TriQXNet's superior forecasting accuracy, setting a new level of expectations for geomagnetic storm prediction and highlighting the potential of classical-quantum hybrid models in space weather forecasting.

Rong Fu, Yiqing Lyu, Chunlei Meng et al.

Automatic generation of radiology reports seeks to reduce clinician workload while improving documentation consistency. Existing methods that adopt encoder-decoder or retrieval-augmented pipelines achieve progress in fluency but remain vulnerable to visual-linguistic biases, factual inconsistency, and lack of explicit multi-hop clinical reasoning. We present NeuroSymb-MRG, a unified framework that integrates NeuroSymbolic abductive reasoning with active uncertainty minimization to produce structured, clinically grounded reports. The system maps image features to probabilistic clinical concepts, composes differentiable logic-based reasoning chains, decodes those chains into templated clauses, and refines the textual output via retrieval and constrained language-model editing. An active sampling loop driven by rule-level uncertainty and diversity guides clinician-in-the-loop adjudication and promptbook refinement. Experiments on standard benchmarks demonstrate consistent improvements in factual consistency and standard language metrics compared to representative baselines.

Moshiur Rahman Tonmoy, Md. Mithun Hossain, Nilanjan Dey et al.

Plant diseases significantly threaten global food security by reducing crop yields and undermining agricultural sustainability. AI-driven automated classification has emerged as a promising solution, with deep learning models demonstrating impressive performance in plant disease identification. However, deploying these models on mobile and edge devices remains challenging due to high computational demands and resource constraints, highlighting the need for lightweight, accurate solutions for accessible smart agriculture systems. To address this, we propose MobilePlantViT, a novel hybrid Vision Transformer (ViT) architecture designed for generalized plant disease classification, which optimizes resource efficiency while maintaining high performance. Extensive experiments across diverse plant disease datasets of varying scales show our model's effectiveness and strong generalizability, achieving test accuracies ranging from 80% to over 99%. Notably, with only 0.69 million parameters, our architecture outperforms the smallest versions of MobileViTv1 and MobileViTv2, despite their higher parameter counts. These results underscore the potential of our approach for real-world, AI-powered automated plant disease classification in sustainable and resource-efficient smart agriculture systems. All codes will be available in the GitHub repository: https://github.com/moshiurtonmoy/MobilePlantViT

Md Abrar Jahin, Md. Akmol Masud, Md Wahiduzzaman Suva et al.

The rapid data surge from the high-luminosity Large Hadron Collider introduces critical computational challenges requiring novel approaches for efficient data processing in particle physics. Quantum machine learning, with its capability to leverage the extensive Hilbert space of quantum hardware, offers a promising solution. However, current quantum graph neural networks (GNNs) lack robustness to noise and are often constrained by fixed symmetry groups, limiting adaptability in complex particle interaction modeling. This paper demonstrates that replacing the Lorentz Group Equivariant Block modules in LorentzNet with a dressed quantum circuit significantly enhances performance despite using nearly 5.5 times fewer parameters. Additionally, quantum circuits effectively replace MLPs by inherently preserving symmetries, with Lorentz symmetry integration ensuring robust handling of relativistic invariance. Our Lorentz-Equivariant Quantum Graph Neural Network (Lorentz-EQGNN) achieved $74.00\%$ test accuracy and an AUC of $87.38\%$ on the Quark-Gluon jet tagging dataset, outperforming the classical and quantum GNNs with a reduced architecture using only 4 qubits. On the Electron-Photon dataset, Lorentz-EQGNN reached $67.00\%$ test accuracy and an AUC of $68.20\%$, demonstrating competitive results with just 800 training samples. Evaluation of our model on generic MNIST and FashionMNIST datasets confirmed Lorentz-EQGNN's efficiency, achieving $88.10\%$ and $74.80\%$ test accuracy, respectively. Ablation studies validated the impact of quantum components on performance, with notable improvements in background rejection rates over classical counterparts. These results highlight Lorentz-EQGNN's potential for immediate applications in noise-resilient jet tagging, event classification, and broader data-scarce HEP tasks.

Md Abrar Jahin, Soudeep Shahriar, M. F. Mridha et al.

Soybean leaf disease detection is critical for agricultural productivity but faces challenges due to visually similar symptoms and limited interpretability in conventional methods. While Convolutional Neural Networks (CNNs) excel in spatial feature extraction, they often neglect inter-image relational dependencies, leading to misclassifications. This paper proposes an interpretable hybrid Sequential CNN-Graph Neural Network (GNN) framework that synergizes MobileNetV2 for localized feature extraction and GraphSAGE for relational modeling. The framework constructs a graph where nodes represent leaf images, with edges defined by cosine similarity-based adjacency matrices and adaptive neighborhood sampling. This design captures fine-grained lesion features and global symptom patterns, addressing inter-class similarity challenges. Cross-modal interpretability is achieved via Grad-CAM and Eigen-CAM visualizations, generating heatmaps to highlight disease-influential regions. Evaluated on a dataset of ten soybean leaf diseases, the model achieves $97.16\%$ accuracy, surpassing standalone CNNs ($\le95.04\%$) and traditional machine learning models ($\le77.05\%$). Ablation studies validate the sequential architecture's superiority over parallel or single-model configurations. With only 2.3 million parameters, the lightweight MobileNetV2-GraphSAGE combination ensures computational efficiency, enabling real-time deployment in resource-constrained environments. The proposed approach bridges the gap between accurate classification and practical applicability, offering a robust, interpretable tool for agricultural diagnostics while advancing CNN-GNN integration in plant pathology research.

Md. Ariful Islam, Md Abrar Jahin, M. F. Mridha et al.

The fast growth of deep learning has brought great progress in AI-based applications. However, these models are often seen as "black boxes," which makes them hard to understand, explain, or trust. Explainable Artificial Intelligence (XAI) tries to make AI decisions clearer so that people can understand how and why the model makes certain choices. Even though many studies have focused on XAI, there is still a lack of standard ways to measure how well these explanation methods work in real-world situations. This study introduces a single evaluation framework for XAI. It uses both numbers and user feedback to check if the explanations are correct, easy to understand, fair, complete, and reliable. The framework focuses on users' needs and different application areas, which helps improve the trust and use of AI in important fields. To fix problems in current evaluation methods, we propose clear steps, including loading data, creating explanations, and fully testing them. We also suggest setting common benchmarks. We show the value of this framework through case studies in healthcare, finance, farming, and self-driving systems. These examples prove that our method can support fair and trustworthy evaluation of XAI methods. This work gives a clear and practical way to improve transparency and trust in AI systems used in the real world.

Md Abrar Jahin, M. F. Mridha, Nilanjan Dey

Feature selection is critical for improving the performance and interpretability of machine learning models, particularly in high-dimensional spaces where complex feature interactions can reduce accuracy and increase computational demands. Existing approaches often rely on static feature subsets or manual intervention, limiting adaptability and scalability. However, dynamic, per-instance feature selection methods and model-specific interpretability in reinforcement learning remain underexplored. This study proposes a human-in-the-loop (HITL) feature selection framework integrated into a Double Deep Q-Network (DDQN) using a Kolmogorov-Arnold Network (KAN). Our novel approach leverages simulated human feedback and stochastic distribution-based sampling, specifically Beta, to iteratively refine feature subsets per data instance, improving flexibility in feature selection. The KAN-DDQN achieved notable test accuracies of 93% on MNIST and 83% on FashionMNIST, outperforming conventional MLP-DDQN models by up to 9%. The KAN-based model provided high interpretability via symbolic representation while using 4 times fewer neurons in the hidden layer than MLPs did. Comparatively, the models without feature selection achieved test accuracies of only 58% on MNIST and 64% on FashionMNIST, highlighting significant gains with our framework. Pruning and visualization further enhanced model transparency by elucidating decision pathways. These findings present a scalable, interpretable solution for feature selection that is suitable for applications requiring real-time, adaptive decision-making with minimal human oversight.

Md Abrar Jahin, Md. Akmol Masud, M. F. Mridha et al.

In high-energy physics, particle jet tagging plays a pivotal role in distinguishing quark from gluon jets using data from collider experiments. While graph-based deep learning methods have advanced this task beyond traditional feature-engineered approaches, the complex data structure and limited labeled samples present ongoing challenges. However, existing contrastive learning (CL) frameworks struggle to leverage rationale-aware augmentations effectively, often lacking supervision signals that guide the extraction of salient features and facing computational efficiency issues such as high parameter counts. In this study, we demonstrate that integrating a quantum rationale generator (QRG) within our proposed Quantum Rationale-aware Graph Contrastive Learning (QRGCL) framework significantly enhances jet discrimination performance, reducing reliance on labeled data and capturing discriminative features. Evaluated on the quark-gluon jet dataset, QRGCL achieves an AUC score of $77.53\%$ while maintaining a compact architecture of only 45 QRG parameters, outperforming classical, quantum, and hybrid GCL and GNN benchmarks. These results highlight QRGCL's potential to advance jet tagging and other complex classification tasks in high-energy physics, where computational efficiency and feature extraction limitations persist.

V. Rajinikanth, Nilanjan Dey, Alex Noel Joseph Raj et al.

Pneumonia is one of the foremost lung diseases and untreated pneumonia will lead to serious threats for all age groups. The proposed work aims to extract and evaluate the Coronavirus disease (COVID-19) caused pneumonia infection in lung using CT scans. We propose an image-assisted system to extract COVID-19 infected sections from lung CT scans (coronal view). It includes following steps: (i) Threshold filter to extract the lung region by eliminating possible artifacts; (ii) Image enhancement using Harmony-Search-Optimization and Otsu thresholding; (iii) Image segmentation to extract infected region(s); and (iv) Region-of-interest (ROI) extraction (features) from binary image to compute level of severity. The features that are extracted from ROI are then employed to identify the pixel ratio between the lung and infection sections to identify infection level of severity. The primary objective of the tool is to assist the pulmonologist not only to detect but also to help plan treatment process. As a consequence, for mass screening processing, it will help prevent diagnostic burden.

Simon James Fong, Gloria Li, Nilanjan Dey et al.

In the advent of the novel coronavirus epidemic since December 2019, governments and authorities have been struggling to make critical decisions under high uncertainty at their best efforts. Composite Monte-Carlo (CMC) simulation is a forecasting method which extrapolates available data which are broken down from multiple correlated/casual micro-data sources into many possible future outcomes by drawing random samples from some probability distributions. For instance, the overall trend and propagation of the infested cases in China are influenced by the temporal-spatial data of the nearby cities around the Wuhan city (where the virus is originated from), in terms of the population density, travel mobility, medical resources such as hospital beds and the timeliness of quarantine control in each city etc. Hence a CMC is reliable only up to the closeness of the underlying statistical distribution of a CMC, that is supposed to represent the behaviour of the future events, and the correctness of the composite data relationships. In this paper, a case study of using CMC that is enhanced by deep learning network and fuzzy rule induction for gaining better stochastic insights about the epidemic development is experimented. Instead of applying simplistic and uniform assumptions for a MC which is a common practice, a deep learning-based CMC is used in conjunction of fuzzy rule induction techniques. As a result, decision makers are benefited from a better fitted MC outputs complemented by min-max rules that foretell about the extreme ranges of future possibilities with respect to the epidemic.

Sourav Samantaa, Nilanjan Dey, Poulami Das et al.

Image Segmentation is a technique of partitioning the original image into some distinct classes. Many possible solutions may be available for segmenting an image into a certain number of classes, each one having different quality of segmentation. In our proposed method, multilevel thresholding technique has been used for image segmentation. A new approach of Cuckoo Search (CS) is used for selection of optimal threshold value. In other words, the algorithm is used to achieve the best solution from the initial random threshold values or solutions and to evaluate the quality of a solution correlation function is used. Finally, MSE and PSNR are measured to understand the segmentation quality.

Arijit Kumar Pal, Poulami Das, Nilanjan Dey

Digital watermarking is a technique of information adding or information hiding in order to identify the owner of the data in multimedia content. It seems that a signal or digital image can permanently embed over another digital data providing a good way to protect intellectual property from illegal replication. The cover data that is transmitted through the internet hides the watermark in a computer aided assertion method such that it becomes undetectable. Finally it stands as a hindrance over many operations without harming the embedded host document. Unfortunately, many owners of the digital materials such as images, text, audio and video are reluctant to the spreading of their documents on the web or other networked environment, because the ease of duplicating digital materials facilitates copyright violation. Digital media distribution occurs through various channels. The cover data may or may not hold any relation with the watermark information. In the last two decades, a considerable amount of research has been done on the digital watermarking of multimedia files such as audio, video, images and text. Different type of watermarking algorithms has been proposed by the researchers to achieve high level of security and authenticity. In our proposed method, a modified reversible watermarking technique is introduced, which employs a blueprint generation of original image based on odd-even embedding methodology to yield large data hiding capacity, security as well as high watermarked quality. The experimental results demonstrate that, no matter how much secret data is embedded, the watermarked quality is about 51dB in this proposed scheme.

Nilanjan Dey, Suvojit Acharjee, Debalina Biswas et al.

In medical field, intravascular ultrasound (IVUS) is a tomographic imaging modality, which can identify the boundaries of different layers of blood vessels. IVUS can detect myocardial infarction (heart attack) that remains ignored and unattended when only angioplasty is done. During the past decade, it became easier for some individuals or groups to copy and transmits digital information without the permission of the owner. For increasing authentication and security of copyrights, digital watermarking, an information hiding technique, was introduced. Achieving watermarking technique with lesser amount of distortion in biomedical data is a challenging task. Watermark can be embedded into an image or in a video. As video data is a huge amount of information, therefore a large storage area is needed which is not feasible. In this case motion vector based video compression is done to reduce size. In this present paper, an Electronic Patient Record (EPR) is embedded as watermark within an IVUS video and then motion vector is calculated. This proposed method proves robustness as the extracted watermark has good PSNR value and less MSE.

Nilanjan Dey, Prasenjit Maji, Poulami Das et al.

In the past few years, like other fields, rapid expansion of digitization and globalization has influenced the medical field as well. For progress of diagnostic results most of the reputed hospitals and diagnostic centres all over the world have started exchanging medical information. In this proposed method, the calculated diagnostic parametric values of the original Electrooculography (EOG) signal are embedded as a watermark by using Difference Expansion (DE) algorithm based reversible watermarking technique. The extracted watermark provides the required parametric values at the recipient end without any post computation of the recovered EOG signal. By computing the parametric values from the recovered signal, the integrity of the extracted watermark can be validated. The time domain features of EOG signal are calculated for the generation of watermark. In the current work, various features are studied and two major features related to blink frequency are used to generate the watermark. The high Signal to Noise Ratio (SNR) and the Bit Error Rate (BER) claim the robustness of the proposed method.

Nilanjan Dey, Pradipti Nandi, Nilanjana Barman

In this paper we propose a method of corner detection for obtaining features which is required to track and recognize objects within a noisy image. Corner detection of noisy images is a challenging task in image processing. Natural images often get corrupted by noise during acquisition and transmission. Though Corner detection of these noisy images does not provide desired results, hence de-noising is required. Adaptive wavelet thresholding approach is applied for the same.

Sayantan Mukhopadhyay, Shouvik Biswas, Anamitra Bardhan Roy et al.

The Electrocardiogram (ECG) is a sensitive diagnostic tool that is used to detect various cardiovascular diseases by measuring and recording the electrical activity of the heart in exquisite detail. A wide range of heart condition is determined by thorough examination of the features of the ECG report. Automatic extraction of time plane features is important for identification of vital cardiac diseases. This paper presents a multi-resolution wavelet transform based system for detection 'P', 'Q', 'R', 'S', 'T' peaks complex from original ECG signal. 'R-R' time lapse is an important minutia of the ECG signal that corresponds to the heartbeat of the concerned person. Abrupt increase in height of the 'R' wave or changes in the measurement of the 'R-R' denote various anomalies of human heart. Similarly 'P-P', 'Q-Q', 'S-S', 'T-T' also corresponds to different anomalies of heart and their peak amplitude also envisages other cardiac diseases. In this proposed method the 'PQRST' peaks are marked and stored over the entire signal and the time interval between two consecutive 'R' peaks and other peaks interval are measured to detect anomalies in behavior of heart, if any. The peaks are achieved by the composition of Daubeheissub bands wavelet of original ECG signal. The accuracy of the 'PQRST' complex detection and interval measurement is achieved up to 100% with high exactitude by processing and thresholding the original ECG signal.

Nilanjan Dey, Pradipti Nandi, Nilanjana Barman et al.

In this paper a comparative study between Moravec and Harris Corner Detection has been done for obtaining features required to track and recognize objects within a noisy image. Corner detection of noisy images is a challenging task in image processing. Natural images often get corrupted by noise during acquisition and transmission. As Corner detection of these noisy images does not provide desired results, hence de-noising is required. Adaptive wavelet thresholding approach is applied for the same.

Nilanjan Dey, Achintya Das, Sheli Sinha Chaudhuri

The present work proposes a computer-aided normal and abnormal heart sound identification based on Discrete Wavelet Transform (DWT), it being useful for tele-diagnosis of heart diseases. Due to the presence of Cumulative Frequency components in the spectrogram, DWT is applied on the spectro-gram up to n level to extract the features from the individual approximation components. One dimensional feature vector is obtained by evaluating the Row Mean of the approximation components of these spectrograms. For this present approach, the set of spectrograms has been considered as the database, rather than raw sound samples. Minimum Euclidean distance is computed between feature vector of the test sample and the feature vectors of the stored samples to identify the heart sound. By applying this algorithm, almost 82% of accuracy was achieved.

Nilanjan Dey, Anamitra Bardhan Roy, Moumita Pal et al.

Segmentation of blood vessels in retinal images provides early diagnosis of diseases like glaucoma, diabetic retinopathy and macular degeneration. Among these diseases occurrence of Glaucoma is most frequent and has serious ocular consequences that can even lead to blindness, if it is not detected early. The clinical criteria for the diagnosis of glaucoma include intraocular pressure measurement, optic nerve head evaluation, retinal nerve fiber layer and visual field defects. This form of blood vessel segmentation helps in early detection for ophthalmic diseases, and potentially reduces the risk of blindness. The low-contrast images at the retina owing to narrow blood vessels of the retina are difficult to extract. These low contrast images are, however useful in revealing certain systemic diseases. Motivated by the goals of improving detection of such vessels, this present work proposes an algorithm for segmentation of blood vessels and compares the results between expert ophthalmologist hand-drawn ground-truths and segmented image(i.e. the output of the present work).Sensitivity, specificity, positive predictive value (PPV), positive likelihood ratio (PLR) and accuracy are used to evaluate overall performance.It is found that this work segments blood vessels successfully with sensitivity, specificity, PPV, PLR and accuracy of 99.62%, 54.66%, 95.08%, 219.72 and 95.03%, respectively.

Tanmay Bhattacharya, Nilanjan Dey, S. R. Bhadra Chaudhuri

This paper proposed a DWT based Steganographic technique. Cover image is decomposed into four sub bands using DWT. Two secret images are embedded within the HL and HH sub bands respectively. During embedding secret images are dispersed within each band using a pseudo random sequence and a Session key. Secret images are extracted using the session key and the size of the images. In this approach the stego image generated is of acceptable level of imperceptibility and distortion compared to the cover image and the overall security is high.

Nilanjan Dey, Moumita Pal, Achintya Das

Digital Retinal Fundus Images helps to detect various ophthalmic diseases by detecting morphological changes in optical cup, optical disc and macula. Present work proposes a method for the authentication of medical images based on Discrete Wavelet Transformation (DWT) and Spread Spectrum. Proper selection of the Non Region of Interest (NROI) for watermarking is crucial, as the area under concern has to be the least required portion conveying any medical information. Proposed method discusses both the selection of least impact area and the blind watermarking technique. Watermark is embedded within the High-High (HH) sub band. During embedding, watermarked image is dispersed within the band using a pseudo random sequence and a Session key. Watermarked image is extracted using the session key and the size of the image. In this approach the generated watermarked image having an acceptable level of imperceptibility and distortion is compared to the Original retinal image based on Peak Signal to Noise Ratio (PSNR) and correlation value.

Tanmay Bhattacharya, Nilanjan Dey, S. R. Bhadra Chaudhuri

This work proposes Steganographic technique for hiding multiple images in a color image based on DWT and DCT. The cover image is decomposed into three separate color planes namely R, G and B. Individual planes are decomposed into subbands using DWT. DCT is applied in HH component of each plane. Secret images are dispersed among the selected DCT coefficients using a pseudo random sequence and a Session key. Secret images are extracted using the session key and the size of the images from the planer decomposed stego image. In this approach the stego image generated is of acceptable level of imperceptibility and distortion compared to the cover image and the overall security is high.

Nilanjan Dey, Anamitra Bardhan Roy, Sayantan Dey

This work proposes a wavelet based Steganographic technique for the color image. The true color cover image and the true color secret image both are decomposed into three separate color planes namely R, G and B. Each plane of the images is decomposed into four sub bands using DWT. Each color plane of the secret image is hidden by alpha blending technique in the corresponding sub bands of the respective color planes of the original image. During embedding, secret image is dispersed within the original image depending upon the alpha value. Extraction of the secret image varies according to the alpha value. In this approach the stego image generated is of acceptable level of imperceptibility and distortion compared to the cover image and the overall security is high.