AI-ASSISTED ART THERAPY THROUGH DIGITAL PLATFORMS Vishakha Akhare, Kapil Mundada, Arivukkodi R, Shikha Gupta, Tejal H. Patil, Balkrishna K Patil Shodhkosh Journal of Visual and Performing Arts, 2026 Digital-based AI-assisted art therapy is a novel discursive field of application of intelligent computing that combines psychological theory, creative practice, and intelligent computing to provide additional mental health services. In the current paper, the author offers a detailed map of how art therapy can be provided through AI-mediated digital space and focuses particularly on emotional expression, reflective interaction, and therapist-mediated intervention. Based on the concepts of expressive and humanistic art therapy, the proposed solution draws on the affective computing and human-computer interaction theories to convert artistic pieces into valuable emotional clues. Machine learning models are used to identify emotion and analyse affect in multimodal inputs, such as drawings, paintings, text stories, and interaction behavior. Guided visual creation is also assisted by generative AI systems, where users are able to express themselves emotionally by means of adaptive prompts, styles and symbolic forms, and reflectively communicate through natural language processing and story generation. They are suggested to have a structured workflow of therapeutic elements that include onboarding, baseline emotional profiling, adaptive creative sessions, and longitudinal visualisation of emotional progress. The application cases show that AI-assisted art therapy has the potential to be used in anxiety, stress management, depression, and neurodiverse users, as well as in remote community-based mental health care.
AI FOR PRESERVING INDIGENOUS FOLK ART PATTERNS Muthukumaran Malarvel Muthukumaran Malarvel, Wamika Goyal, Sadhana Sargam, Kapil Mundada, A. Viji Amutha Mary, Ashwika Rathore Shodhkosh Journal of Visual and Performing Arts, 2025 The indigenous folk art traditions represent centuries of cultural wisdom, community belonging, symbolism, and craftsmanship which are specific to a certain region. The continuity of these visual heritage systems is however endangered by fast urbanization, erosion of artisanal transmission and little digital documentation. The presented paper is an AI-based model of the preservation of indigenous folk art patterns by creating systems of data, developing features, and using motifs to reconstruct the object. The suggested approach combines controlled imaging pipelines based on museums, archives, craftsmen and field-based surveys to come up with a very enriched, metadata-conformant, cultural heritage information dataset. Annotation protocols represent the geometry of the motifs, stroke behavior, chromatic palettes, material references and regional semantic attributes that allow the structured cultural representation. The system identifies the discriminative visual features, which are used to define the various folk art traditions, using deep learning models DNNs, ViTs and multimodal style-embedding models. Moreover, encoding symbolic motifs allows comparing styles crossly, and clustering patterns as well as synthesizing them using AI-like synthesis models. Experimental findings indicate that there is a high performance in the motif recognition accuracy, reconstruction fidelity and stylistic consistency indicating a possibility of AI to support artisans, researchers and cultural institutes. The discussion showcases the importance of the ethical aspects of the community involvement, cultural sensitivity and responsible practices of digitization, not just those related to the technical developments. On the whole, the current work adds a scalable, culturally sensitive AI system with the purpose of preserving the indigenous visual knowledge and guaranteeing its continued transfer to the subsequent generations.
AI-DRIVEN MUSIC CURATION AND VISUAL CULTURE: AUDIENCE PREFERENCE ANALYSIS IN DIGITAL CREATIVE PLATFORMS Utkarsh Verma, Apurba Chakraborty, Kapil Mundada, Jay Vasani, Nitin Rakesh, Manisha Vilas Khadse Shodhkosh Journal of Visual and Performing Arts, 2025 Social music curation has turned into a hallmark of AI-powered creative platforms defining the way digital platform forms listening habits and the visual culture, as well as the way to engage with the audience. This paper discusses the connection between AI and music recommendation which applies visual aesthetics to analyse the audience preferences in a modern digital creative ecosystem. The issue being discussed involves the lack of transparency related to the formation of cultural preferences by algorithms and the insufficient knowledge about the impact of music -visual associations on the perception and retention of users. This study aims to simulate the trend of audience preference through the combined analysis of an audio characteristic, visual representation, and behavioral interaction information. The suggested approach utilizes multimodal deep learning, which is audio embedding networks, visual feature extractions models and attention-based preference learning to elicit cross-modal associations among sound, images, as well as user reaction. Data of large scale interaction on digital platforms is treated to recognize clusters of preferences, and changes in taste over time, and the interaction-based visual-music fit. The results of the experiments suggest that AI-based curation greatly improves the audience satisfaction, and the curation produces some measurable results in terms of the engagement time, the variety of its discoveries, and the subjective aesthetic cohesion. Findings also demonstrate that contextualized music recommendations House better than audio only systems in predicting user preferences and maintaining creative exploration.
THE ROLE OF AI IN DEMOCRATIZING VISUAL STORYTELLING Bijal Jigar Talati, Garima Jain, Manvinder Brar, K Prakash, Veerendra Yadav, J P Yadav, Kapil Mundada Shodhkosh Journal of Visual and Performing Arts, 2025 The fast development of artificial intelligence turned visual storytelling into a more approachable, inclusive and participatory creative action. Historically, visual narratives were only done well by specialized artistic skills and costly equipment, and a barrier of experience in production that made visual media communication only available to a few. This can be minimized by AI-enhanced technologies like generative image models, style-transfer systems, multimodal storytelling systems, and automated editing pipelines which allow users with diverse backgrounds to tell captivating visual stories with more ease than ever before. They favor ideation, the structuring of narrative, composition of a scene, and aesthetic polishing, allowing individuals with no formal training in art to visualize themselves and experiment in a creative way. Also, since AI personalises visual content based on cultural and emotional subtexts and user-specific intentions, it creates more genuine and meaningful storytelling. It also facilitates usability with voice-to-image synthesis, user-friendly interfaces to creators with limited abilities and language-neutral narrative editors. AI promotes creative confidence in learning settings, which allows students to cycle quickly and visualize abstract ideas. On a societal level, AI democratizes the process by enabling marginalized groups to get new channels to conserve narratives, rebrand traditions, and recount lived experiences on the digital realms. Along with these developments, such aspects of ethics as authorship, originality, bias, and responsible deployment are also of critical importance. On the whole, AI is a radical power that broadens the range of creators, the format of stories, and those whose voices are heard within the visual storytelling system.
INTEGRATING AI-ART TOOLS INTO FINE ARTS CURRICULA Hemalatha BS, Swati Srivastava, Rama Choudhary, Ramesh Saini, Kanika Seth, Kapil Mundada Shodhkosh Journal of Visual and Performing Arts, 2025 The introduction of AI-art tools into the curriculum of the field of fine arts is a drastic deviation of the current art education process, as it alters how students perceive, experiment, and construct a work of visual art. As more and more generative algorithms, diffusion models, and interactive AI systems are made accessible, they offer novel forms of creativity through more traditional forms of media and processes. The paper explains how AI-art tools can be incorporated into the fine art course in an intellectual manner to enhance the creative inquiry, diversify visual problem solving, and interdisciplinary skills that will be used in the artistic used sectors in the future. In accordance with the sources related to the theme of digital art education and current tendencies in the field of creative technologies based on AI, the paper will provide the pedagogical premises of the curricular integration, the importance of shaping adaptability, critical thinking, and technological fluency in art students. It is proposed to use a hybrid creative process, and the design of a framework of integration, specific curriculum modules, studio practices, centered around AI processes and the workflow, and process evaluation in terms of AI processes and approaches. The framework does not claim that AI is going to replace the fundamental artistic abilities but, instead, it will act as an extension that will enable the individuals to discover more, develop quicker, and democratize numerous sophisticated creative means. The final product of the AI-enhanced art education may be positive with more developed imagery, cross-disciplinary interaction, and keeping up with the latest tendencies in the sphere.
Quantum Entanglement Dynamics in Topological Insulators for Next-Generation Quantum Computing Applications Melam Thirupathaiah, Vinayak Vasantrao Mukkawar, J. Mary Praveena, Prakhar Goyal, Kamineni Sairam, Kapil Mundada, Baoxin Le International Academic Journal of Science and Engineering, 2025 The stability of quantum entanglement has been a bottleneck in the design of scalable quantum computing architectures, especially due to decoherence and environmental noise in traditional quantum materials. Topological insulators (TIs), which have bulk-insulating states and conducting surface states protected by time-reversal symmetry, offer a promising platform for enhancing entanglement resilience. This paper explores the mechanism of quantum entanglement in two-dimensional topological insulator systems and assesses their potential for next-generation quantum computing. -Using tight-binding Hamiltonian simulations and time-dependent density-matrix simulations, we study the evolution of entanglement as a function of temperature, disorder strength, spin-orbit coupling, etc. The entanglement is measured using concurrence and von Neumann entropy measures among a variety of system setups. Simulations Findings TI-based qubit systems have been found to have a 35-48 percent improvement in entanglement lifetime over conventional semiconductor qubits with comparable noise levels. Remarkably, coherence times of over 120 m are observed at temperatures lower than 50 mK, and an estimated 42% slows down the rate of entanglement degradation in the presence of moderate non-magnetic disorder. Moreover, the statistical analysis of 1,000 simulation runs shows that the standard deviation of entanglement fluctuations has decreased by a factor of 31, leading to better stability and reproducibility. These data demonstrate the inherent fault-tolerant properties of the topological surface states and their ability to maintain quantum correlations over longer periods of operation. Finally, the conclusions indicate that topological insulators provide a viable, scalable material platform for maintaining quantum entanglement, with important implications for quantum logic models and error-tolerant quantum computing. This paper justifies the incorporation of topological materials into future quantum information processing systems.
Predictive Heart Rate Monitoring System Kapil Mundada, Utkarsha Kavitake, Dhruva Mankar, Sumedh Jadhav, Shivesh Marhatta 2025 6th International Conference for Emerging Technology Incet 2025, 2025 The increased frequency of long-term sickness along with the widespread impact of cardiovascular diseases as a primary cause of death have become a common concern nowadays and emphasize the necessity for easily accessible, real-time health monitoring and early detection. There is a specific product in front of us that we are going to talk about now which is about an advanced set of hardware components and algorithms to give the user an opportunity to predict their own heart health. This system is manufactured in the form of an AmazingMAX30100 sensor and NodeMCU microcontroller that are taken as physical entities that are measuring the patient's health proportions like heart rate and oxygen saturation (SpO₂) which in turn are shown on an OLED screen in real-time for the patient and at the same time are sent to a cloud platform for the physician to follow on remotely. A data science-driven software further strengthens the proposed system, by employing the latest techniques in machine learning such as Random Forest, Decision Tree, Support Vector Machine, and Logistic Regression to seek out the sources of heart disease, if any. It goes beyond this by also embodying a simple web application to give users the power of full interaction with the system and with each stage, i.e. inputting their own data into the system, accessing their health parameters, getting real-time information and accessing their individual risk.
Automated X-Ray Image Stitching for Enhanced Medical Diagnostics and Visualization Kapil Mundada, Shivam Dapkekar, Shambhavi Deshpande, Raghav Deshpande, Sweety Deshmane 2025 2nd International Conference on Integration of Computational Intelligent System Icicis 2025, 2025 Medical image stitching moves past segmental projection in current medical imaging as it integrates multiple X-ray or CT images into a full image. Using stacking images is essential for enabling better diagnostic accuracy in orthopedics, dentistry, trauma, and oncology, as there is a great need to visualize complex diagnosis and treatment. While stitching offers up a more accurate image with fewer misregistration errors compared to other pan radiography composition techniques, stitching is also prone to errors of misalignment, which may simulate fractures, seams, or ghosting artifacts. Current techniques using stitching is mostly affected by not enough exposures, misaligned shooting, or object geometries or shapes. This project is built upon the features of extracting images, estimating homographies, and image blending method that should automate the X-ray image stitching project in Python using OpenCV. The proof of concept tool will hopefully support and minimize repeated imaging (and thereby patient exposure to radiation) while being easily/capably integrated into existing radiologist workflows in less time. In addition, it will provide an accurate and coherent visualization to assist in advancing patient care in the context of emergency, oncology, or veterinary hospital contexts for clinical decision making.
Customer Churn Prediction in E-Commerce Using AutoML and Explainable AI Frameworks Shiney Chib, Kapil Mundada, Kanika Handa, Abhishek Vishwakarma, Keerthi Jain, R. Ragul Kannan 2025 International Conference on Emerging Trends in Networks and Computer Communications Etncc 2025 Proceedings, 2025 E-commerce companies that want to keep their customers and make as much money as possible need to be able to guess when customers will leave. With the fast growth of data-driven methods, businesses can now correctly predict which customers will leave, letting them take preventative steps. This study looks into how Automated Machine Learning (AutoML) and Explainable AI (XAI) systems can be used to predict customer loss in online shopping. AutoML is used to handle model selection, feature engineering, and hyperparameter tuning. This improves the accuracy of predictions without the need for human input. On the other hand, XAI is added to make models more clear and easy to understand so that everyone involved in the business can trust and understand the decision-making process. Utilising a number of supervised learning models, such as Random Forest, Gradient Boosting, and Neural Networks, the study tests model performance and finds the best ways to predict loss. Adding XAI tools like SHAP (Shapley Additive Explanations) and LIME (Local Interpretable Model-Agnostic Explanations) also helps understand the reasons behind loss predictions, which makes it easier for decision-makers to create strategies that keep customers. The results show that the suggested AutoML-XAI method not only makes accurate predictions, but also gives useful information about how customers act. It's becoming more and more important for e-commerce businesses to have automated solutions that can be explained. This study helps businesses make better decisions and keep customers.
Early Fire Detection Using Deep Learning Akshad Jha, Saurabh Vedak, Kapil Mundada, Raj Walnuskar, Utkarsh Chopade, Anand Iyer Proceedings 2021 1st IEEE International Conference on Artificial Intelligence and Machine Vision Aimv 2021, 2021
MR image restoration using non-linear filtering Vedant R Shukla, Kapil Mundada, Jayant Kulkarni 2018 3rd IEEE International Conference on Recent Trends in Electronics Information and Communication Technology Rteict 2018 Proceedings, 2018
Automatic cleaning mechanism for rain guage Shalaka D. Urkude, Kapil G. Mundada Rteict 2017 2nd IEEE International Conference on Recent Trends in Electronics Information and Communication Technology Proceedings, 2017
