The DSpace digital repository system captures, stores, indexes, preserves, and distributes digital research material. https://repository.iiitd.edu.in:443/xmlui 2026-09-14T06:23:46Z 2026-09-14T06:23:46Z Sarkar, Neelanjan Yadav, Shubham Chakravarty, Sambuddho (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/2133 2026-09-12T22:00:43Z 2024-11-27T00:00:00Z

Secure VoIP communication over tor Sarkar, Neelanjan; Yadav, Shubham; Chakravarty, Sambuddho (Advisor) This report presents the implementation, analysis, and revaluation of secure VoIP communica tion using the Tor network. The primary objective of this research is to revaluate the feasibility of deploying VoIP systems over Tor while maintaining adequate performance, minimal latency, and high-quality call output. In an era where data privacy and security are increasingly im portant, this study explores how integrating VoIP with Tor can provide a robust, anonymous communication solution for privacy-conscious users. The project integrates OpenVPN with Tor to enhance data security, leveraging Tor’s distributed relay network to anonymize traffic and obscure communication metadata. This approach ad dresses growing concerns about mass surveillance, data breaches, and cyber threats, ensuring that sensitive voice communications remain protected. Furthermore, the report delves into the unique challenges posed by Tor’s architecture, such as increased latency, jitter, and packet loss, which are detrimental to real-time applications like VoIP. A comprehensive evaluation methodology has been developed, which includes the measurement of call quality using PESQ (Perceptual Evaluation of Speech Quality) and the analysis of critical network metrics such as latency, jitter, and routing configurations. The experimental setup involved large number of iterations of VoIP calls routed through dynamically configured Tor circuits, enabling a detailed assessment of performance across varying conditions. The results of this study demonstrate that, under optimized configurations, VoIP calls over Tor can achieve acceptable quality levels. The research highlights that careful selection of Tor nodes, optimized routing strategies, and a balance between privacy and performance are key to achieving reliable communication. This work underscores the potential of Tor as a tool for real-time secure communication, showcasing its applicability beyond traditional web traffic anonymization. Additionally, the study provides insights into potential optimizations for reducing Tor-induced latency and improving call quality. By addressing these challenges, this research contributes to the growing field of secure communication technologies and highlights the trade-offs between user anonymity and performance in real-time systems. This report aims to serve as a foundation for future explorations of privacy-centric VoIP systems, offering a pathway for further innovation in this domain.

2024-11-27T00:00:00Z Parikh, Hardi Mehroliya, Shlok Singh, Pushpendra (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/2132 2026-09-12T22:00:26Z 2024-11-27T00:00:00Z

Full stack development- kushalma Parikh, Hardi; Mehroliya, Shlok; Singh, Pushpendra (Advisor) The objective of our website, KushalMa, is to provide invaluable support to expectant mothers residing in rural locales. We aim to deliver this through an established network of moderators and system administrators. Our service is intricately designed to process and analyze interactions from WhatsApp chats and Zoom calls. These communications are collected from a diverse array of groups strategically located throughout India, with a particular emphasis on the regions of Punjab, Maharashtra, and Madhya Pradesh. Our platform facilitates structured assistance, enabling administrators to seamlessly upload WhatsApp transcripts for comprehensive analysis. By leveraging these insights, we can generate and present meaningful statistics on our website. This initiative is part of our commitment to enhancing the care and guidance available to mothers during this critical phase of life.

2024-11-27T00:00:00Z Barala, Avinash Khan, Saad Bagler, Ganesh (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/2131 2026-09-12T22:00:38Z 2024-11-27T00:00:00Z

Computational gastronomy: web development and database creation Barala, Avinash; Khan, Saad; Bagler, Ganesh (Advisor) In this project, we have developed a comprehensive database and web platform for multiclass toxin protiens and peptides by integrating data from trusted and reputed sources such as Arach noServer and UniProt. The objective of our project is to provide easy access to efficiently predict the toxicity of the peptide using Sequence to the facilitating research and development in related fields. we started from collecting and classifying the details information about various toxin pro tein,we gathered over 7,000 data entries, which were then filtered and classified into five main categories: Neurotoxin, Cytotoxin, Hemotoxin, Enterotoxin, and Cardiotoxin. We used the MERN (MongoDB, Express.js, React.js, Node.js) stack, we created a dynamic website that al lows users to search and retrieve toxin information based on various criteria. We integrated a machine learning ensemble model with an accuracy of 91.05% on training data and 87.36% on testing data to predict the toxicity of peptide sequences. This tool enhances the platform’s utility by providing predictive insights into sequences not already present in the database. This Implimentation holds a significant resource for researchers and advance the understanding of toxin peptide.

2024-11-27T00:00:00Z Prasad, Tejash Singhal, Tanmay Garg, Madhav Krishan Kumar, Dhruv (Advisor) http://repository.iiitd.edu.in/xmlui/handle/123456789/2130 2026-09-12T22:00:32Z 2024-11-27T00:00:00Z

LLM-based research assistant Prasad, Tejash; Singhal, Tanmay; Garg, Madhav Krishan; Kumar, Dhruv (Advisor) This project explores the capabilities of Large Language Models (LLMs) as research assistants by developing an advanced research paper reviewer that surpasses current leading tools such as AI-Scientist developed by SakanaAI. Initially, using foundational APIs from Google Gemini and OpenAI’s GPT models, we created a baseline reviewer that generated generic feedback in adherence to target conference guidelines. To enhance its effectiveness, we implemented a sophisticated pipeline incorporating agentic patterns and reflection-based models, enabling iterative refinement of reviews for increased specificity and accuracy. Using ExtractorAPI, the system retrieves conference-specific review guidelines, allowing the LLM to dynamically split research papers into independent sections for detailed evaluation. In subsequent iterations, we transitioned to a LangGraph-based architecture, adopting a multi agent approach with a supervisor node and integrated tools like internet search and Semantic Scholar to emulate a professional reviewer’s comprehensive capabilities. This enhanced workflow not only automates the review process but also delivers high-quality, conference-aligned feedback more efficiently than existing solutions. Our results demonstrate the potential of LLMs to significantly streamline the research review process, offering scalable and consistent support to researchers.

2024-11-27T00:00:00Z