arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Author

Deepa Parasar

14 papers · 2015–2026

2015: 32016: 12017: 02018: 02019: 02020: 02021: 22022: 12023: 02024: 12025: 02026: 6
2015peak 6 in 20262026

Frequent co-authors

Papers

showing 1–14 of 14 RSS

Sailfish invasive weed optimization algorithm for multiple image sharing in cloud computing

Vitthal Gutte, Deepa Parasar

Computer Networks and Communications · International Journal of Intelligent Systems · · cited 18× · DOI ↗ · more like this ·

Visual cryptography (VC) is a secret sharing cryptographic technique, which permits visual information that is encrypted in such a way that the decrypted information shows the visual image. Security is the common problem that arises in the general VC technique. To secure the visual information, an effective method is proposed using a developed sailfish invasive weed optimization (SIWO) algorithm. At first, the grayscale image is shared with multiple parties, and here the shared images are encrypted and decrypted to reconstruct the original image back. The key generation is done using the developed SIWO-based data sharing approach. Here, the developed approach offered the peak signal-to-noise ratiof 34.870 and 33.711 dB for Lena and Cameraman images, respectively, and it attains better performance in terms of secret image sharing.

Seeded region growing segmentation on ultrasound image using particle swarm optimization

Parineeta Suman, Deepa Parasar, Vijay R. Rathod

Computer Vision and Pattern Recognition · · cited 7× · DOI ↗ · more like this ·

Ultrasound imaging is one of the most popular and cheapest noninvasive medical scans. At the time of image acquisition, there may be degradation in the quality of image in the form of speckle noise. In recent times, many researches have made various experiments to enhance the quality of medical imaging. However, there is scope to further enhance it. In the proposed method, finding out the seed pixel randomly is the basic problem, which is treated as an optimization problem. This problem can be solved by Particle Swarm Optimization. Using Particle Swarm Optimization algorithm, the fitness function can give us the appropriate seed pixel for the desired ultrasound imaging. In this paper, a novel method is proposed, wherein segmentation will be applied on a fuzzy filtered image. The fuzzy filter applies fuzzy rules to detect regions in the image viz. edge region, homogeneous region, and…

Recommendation system using machine learning

Suhasini Parvatikar, Deepa Parasar

Artificial Intelligence · International Journal of Artificial Intelligence and Machine Learning · · cited 6× · DOI ↗ · more like this ·

Due to increase of data on internet, there is an increased dependency on internet by people Thus, recommendation systems help people by suggesting products where is overload of information on ecommerce websites. There are various methods for recommendation. This paper study about various techniques used in designing of recommendation system with machine learning algorithm.

A COMPARATIVE ANALYSIS FOR STOCK PRICE PREDICTION USING IMPROVED EXPANSIVE DEEP LSTM MODEL

Pankaj Rambhau Patil, Deepa Parasar, Shrikant Charhate

Artificial Intelligence · Indian Journal of Computer Science and Engineering · · cited 3× · DOI ↗ · more like this ·

Due to a lack of clarity and flexibility, prediction leveraging ML models is not well fitted in many sections of commercial decision processes. Proposed model aim to employ deep learning strategy in the stock market pricing area to generate positive risk-adjusted price by analyzing previous transaction data and maintaining greater accuracy with a lower error rate. In this study, the deep learning approach is used, which is capable of handling time-series data. The results are obtained with evaluation of error rate metric MSE & RMSE which express how distant the data points are from the regression line. RMSE measures the dispersion of these residuals. It shows how concentrated the data is on the best fit line. This study compares a unique deep learning methodology with deep LSTM, GA and Harris Hawk optimization. As a part of this analysis results are observed and plotted for the various…

Spotlight : Redefining Theatre Application Through AI-Driven Personalization and Community Engagement

Deepa Parasar, Swetta Kukreja, Vedang Vaidya +3 more

Artificial Intelligence · · cited 1× · DOI ↗ · more like this ·

Spotlight, a platform dedicated to the dramatic arts, aims to change spotlight at theatres and redefine the experience at theatres. This research paper aims to depict the evolution of Spotlight by emphasizing on its usage of Kotlin and Android Studio, making use of which it provides user-friendly UI. Spotlight is not only a mobile application, but it uses artificial intelligence (AI) which can generate personalized scripts automatically that are based on preferences, thus, enhancing user experience. Users, by recording and sharing their performances of the scripts, facilitate the community involvement and self-expression channel. What gives Spotlight the main benefit is their interactive community feature, which highlights ratings and supports everyone needs to create a feeling of belonging. This paper illustrates the engineering details of Kotlin, Android Studio, as well as AI…

XAI Decoded

Swetta Kukreja, Deepa Parasar, Alesha Ali +1 more

Artificial Intelligence · CRC Press eBooks · · cited 0× · DOI ↗ · more like this ·

Explainable artificial intelligence (XAI) aims to close that gap by producing human-understandable outputs that establish trust, transparency, and accountability between people and AI systems. This chapter provides a systematic literature review (SLR), according to PRISMA, to analyze peer-reviewed studies published between 2015 and 2024. The review also differentiates between local and global explanation methods in terms of whether they concentrate on individual predictions or on general model behavior. Based on our findings, SHAP and LIME are overwhelmingly used in practice because they work with black-box models and are easy to implement, although the two methods differ in their theoretical foundation and computational complexity. Despite the increasing relevance of XAI in safety-critical areas such as healthcare, finance, and autonomous systems, open issues remain concerning the…

Research challenges and future trends in Blockchain of Things

Shraddha C. Subhedar, Deepa Parasar

Information Systems · · cited 0× · DOI ↗ · more like this ·

Blockchain of Things (BCoT) refers to the alliance of two thriving technologies: Blockchain and Internet of Things (IoT). Although IoT has given us new ways to connect with our surroundings, it incurs concerns about privacy, security, and cost as it exploits functionality of centralized architecture. International Data Corporation (IDC) calculated that 41 billion IoT devices connected to the internet will be generating 79.4 zettabytes of data by 2025. In traditional cloud-based IoT, structured data is composed by a centralized server which causes a single point of failure. Additionally, some of the components of the IoT network might demand high data privacy, huge data storage with quick response time. These demands can be taken as probing challenges for the whole IoT network. BC can be defined as a distributed append-only public ledger technology that was originally proposed for…

Efficient Gun and Knife Detection Utilizing the YOLOv5 Model

Swetta Kukreja, Deepa Parasar, Saad Shaikh +3 more

Computer Vision and Pattern Recognition · · cited 0× · DOI ↗ · more like this ·

Early and proper identification of weapons such as guns and knives is a key element in enhancing security in several locations. In this paper, we made a massive experiment with the YOLOv5 model to identify the gun and knife in the photo and video. We selected a deep learning architecture named YOLOv5 due to its speed/accuracy trade-off, which is most suitable in real-time localizations in video streams and images. The data is available in Google images via web-scraping and a fair portion of the images via manual collection to train. To make our model more efficient and more accurate, we annotated the images individually by hand with the utmost human level of accuracy under various lighting conditions. The authors concluded that YOLOv5 could be used successfully and that its ability to detect motions is likely to become a significant weapon test tool.

A fully homomorphic encryption federated learning architecture for privacy preserving in industrial internet of things

Shraddha C. Subhedar, Dr. Deepa Parasar

Artificial Intelligence · MethodsX · · cited 0× · DOI ↗ · more like this ·

We are currently entering the fifth revolution of industry - Industry 5.0. IIoT is the domain where massive quantities of data are flourished by the associated devices in an industry on a daily basis. To realize industry 4.0, the Industrial internet of things is considered a prominent one. Federated Learning, also known as collaborative learning, employs a decentralized approach in its applicability while maintaining data privacy, but many existing frameworks struggle with handling privacy of gradients which are transferred to Federated servers. Unlike conventional approaches, proposed fully homomorphic encryption based Federated Learning-FHEEFL ensures that raw gradients never leave local IoT nodes; instead, only updates or changes in model secured with encryption techniques are transmitted. The framework achieves a very strong privacy as well as data security. FHEEFL is intended for…

Advanced multimodal multitask framework for cyberbullying detection

Devyaneeraje Deshmukh, Deepa Parasar, Swetta Kukreja +1 more

Artificial Intelligence · IET conference proceedings. · · cited 0× · DOI ↗ · more like this ·

The detection of cyberbullying in digital spaces becomes difficult because of how sarcasm and implicit emotions and multiple communication signals from text and audio and video content interact with each other. The current text-based models demonstrate poor performance in detecting both implicit and subtle abuse because they lack the ability to understand contextual and affective information. The research introduces an enhanced multimodal multitask learning system which uses sentiment analysis and emotion recognition and sarcasm detection as supporting tasks to boost cyberbullying detection accuracy. The framework combines pretrained models BERT for text analysis with Wav2Vec2 for audio processing and R3D-18 for video analysis through attention-based fusion and cosine similarity-based alignment. The system uses 120 annotated samples from four different media types to train and test its…

Enhancing mineral exploration using hyperspectral imaging and dual regression reflectance conversion

Deepa Parasar, Swetta Kukreja, Drashya Sodha +3 more

Artificial Intelligence · IET conference proceedings. · · cited 0× · DOI ↗ · more like this ·

This paper proposes a novel, automated system for mineral detection using hyperspectral imaging (HSI), with a focus on kaolinite identification. Unlike traditional approaches, our method introduces a dual regression model that learns to convert top-of-atmosphere (TOA) reflectance to bottom-of-atmosphere (BOA) reflectance, treating atmospheric correction as a denoising problem. The model is trained on paired TOA-BOA reflectance data and supports bidirectional learning for improved stability and spectral fidelity. Post-correction, spectral angle mapping is used to identify mineral presence from spectral libraries. The experimental results from EnMAP satellite imagery show both high accuracy and spatial reliability which validates the model's performance. The research combines atmospheric correction with mineral classification into a complete system that supports real-time and scalable…

Comparison of Fuzzy Filter and Median Filter on Ultrasound Image

Parineeta Suman, Deepa Parasar, Vijay R. Rathod

Computer Vision and Pattern Recognition · · cited 0× · OpenAlex ↗ · more like this ·

Ultrasound imaging is one of the popular,cheapest, and non-invasive medical scans. During the time of image acquisition, there is distortion of the quality of image in the form of speckle noise. Nowadays, many researches have made various experiments to improve the quality of medical image.In this paper, comparing fuzzy filter with median filter, as fuzzy filter works on fuzzy rules to detect different gradient and then filtering the noisy, homogenous and edge regions whereas median filter works on the pixel value in a non-linear fashion.The performance comparison of multi-scale schemes shows that fuzzy filter excels over the median filter.

Reversible color transmission of compressed fragment-visible mosaic image

Smita Kuldiwar, Deepa Parasar

Computer Vision and Pattern Recognition · · cited 0× · DOI ↗ · more like this ·

This paper presents Reversible Color Transmission of Compressed Fragment -Visible Mosaic Image of the same size for hiding the secret image. By applying the digital watermarking on mosaic image using DWT and PCA transformation increase the security, authentication, copyright protection and all types of internet applications. Most commonly used methods of watermarking are discrete cosine transform (DCT), discrete wavelet transform (DWT).A hybrid digital image watermarking scheme based on Discrete Wavelet Transform (DWT) and Principal Component Analysis (PCA). These transform domain technique provides criteria of imperceptibility as well as robustness than Discrete Cosine Transform and Discrete Wavelet Transform. The proposed method, originally designed for dealing with color images. An additional approach here is used to enhance the embedded watermark data security is also proposed. The…

Study on the Reduction of Speckle Noise in Ultrasound Images

Parineeta Suman, Deepa Parasar, Vijay R. Rathod

Computer Vision and Pattern Recognition · · cited 0× · OpenAlex ↗ · more like this ·

The ultrasound images, the speckle noise is inherent in medical ultrasound images and it is the cause of reduced contrast-to-noise ratio and resolution. The presence of speckle noise is not attractive because it reduces the image quality and affects the tasks of the individual interpretation and paper diagnosis. In this paper, we study various techniques to reduce speckle noise from ultrasound images Post acquisition method as the only scale spatial filtering method and multi-scale method.