Research

Ch Nikhilesh Krishna, Avishek Rauniyar, N Kireeti Sai Bharadwaj, Sujay Bharath Raj, Vipina Valsan, Kavya Suresh, V Ravikumar Pandi, Soumya Sathyan

International Conference on Information and Communication Technology for Smart Systems (ICTCS 2023)

Forest ecosystems face increasing threats from climate change, illegal logging, and human encroachment. Traditional wildlife monitoring methods are labor-intensive, expensive, and fail to provide real-time insights for conservation decision-making. ECO-Guard presents an integrated AI sensor system that combines computer vision, IoT sensor networks, and edge computing for continuous wildlife monitoring and sustainable forest management. The system deploys camera traps with on-device ML inference for species detection, environmental sensors for habitat monitoring, and a centralized dashboard for conservation analytics.

S Adithya Krishna, R Sreeram, Avishek Rauniyar, BM Reddy, JD Udayan

International Conference on Information and Communication Technology for Competitive Strategies (ICTCS 2024)

Electronic voting systems promise increased accessibility and efficiency in democratic processes, but must balance security, usability, and verifiability. PyVote presents a secure and efficient electronic voting system featuring a user-friendly graphical user interface built in Python. The system implements cryptographic protocols for ballot security, voter authentication, and audit trail generation, while maintaining an intuitive interface accessible to non-technical voters.

S Adithya Krishna, R Sreeram, Avishek Rauniyar, BM Reddy, S Sarath

International Conference on Information and Communication Technology for Competitive Strategies (ICTCS 2024)

A systematic comparative analysis of machine learning classification models across diverse datasets and application domains. The chapter examines algorithm performance through the lens of accuracy, computational efficiency, interpretability, and generalization capability. By presenting a structured comparison framework across multiple dataset types, the work helps practitioners make informed algorithm selection decisions based on problem characteristics rather than popularity or convenience.

Avishek Rauniyar, Bowrampet Manish Reddy, Remya S Dr.

Procedia Computer Science

Prostate cancer remains one of the most prevalent cancers globally, with treatment outcomes heavily dependent on accurate staging and personalized treatment planning. This paper presents a comprehensive deep learning framework for stratifying prostate cancer patients across multiple clinical dimensions — Gleason grading, TNM staging, and treatment pathway classification. The framework integrates histopathological imaging features with clinical biomarkers to produce multi-level risk assessments. Published in Procedia Computer Science Volume 259, pages 356–365.