Latest Updates
To know the details for B.Tech. (Hons.) seat allotment schedule under JEE Main Quota, please click here | Ph.D. Admissions 2026 Now Open | Full-Time and Part-Time | Apply Before 10 April 2026 – Apply Now | Information submitted to UGC for inspection purpose | UGC Public Self Disclosure | CET Code: E285 | CET Code: E285 / MBA – CET Code – B413 / M.Tech – CET Code – T615

Aashish Khilnani

Aashish Khilnani

Assistant Professor

About

Aashish Khilnani has completed his PhD from Banaras Hindu University. His doctoral thesis, titled “A Study on Relevant Feature Ensemble Techniques for EEG-based Brain-Computer Interfaces.” focused on developing relevant and effective computational approaches for Brain-Computer Interfaces (BCIs). His research interests include the analysis of Electroencephalogram (EEG) signals, Brain-Computer Interfaces, Explainable Artificial Intelligence, Graph Neural Networks, Adversarial Attacks, Signal Processing, Machine Learning and Deep Learning.

He has an academic background in Mathematics, having completed his B.Sc. (Hons.) in Mathematics from Jamia Millia Islamia and M.Sc. in Applied Mathematics from South Asian University. He subsequently moved towards Computer Science and pursued his PhD at the Centre for Interdisciplinary Mathematical Sciences, Banaras Hindu University. This interdisciplinary academic journey has provided him with a foundation in both Mathematics and Computer Science. He has qualified for the CSIR-UGC NET in Mathematical Sciences and the UGC NET in Computer Science and Applications. He has also qualified for the GATE examination in Data Science and Artificial Intelligence.

In addition to his academic and research interests, he enjoys playing chess and has a keen interest in long-distance running and marathon training.

"If our brains were simple enough for us to understand them, we'd be so simple that we couldn't." - Ian Stewart

Enhancing Adversarial Attack Detection in EEG Signals With Covariance Entropy: A Novel Framework for BCI Security
Published in IEEE Signal Processing Letters. This work introduces Covariance Entropy (CovEn), a novel entropy-based approach for detecting noise and adversarial attacks in EEG signals. Unlike conventional entropy measures that primarily analyse individual signal characteristics, CovEn incorporates covariance matrices to capture both the variability and inter-channel relationships present in multichannel EEG signals. The experimental results demonstrate that CovEn is sensitive to signal contamination and can effectively distinguish clean EEG signals from poisoned or attacked signals. This work contributes towards improving the security, robustness, and reliability of EEG-based BCI systems.

Link: doi.org/10.1109/LSP.2025.3580330

A Dual Adaptation Approach for EEG-Based Biometric Authentication Using the Ensemble of Riemannian Geometry and NSGA-II
Published in International Conference on Pattern Recognition. This work presents a methodology for EEG-based biometric authentication that combines Riemannian geometry with multi-objective optimisation. EEG signals are represented by symmetric positive-definite covariance matrices. These covariance matrices are then analysed using a Riemannian geometry framework. Moreover, to address the high dimensionality and computational cost associated with multichannel EEG signals, the Non-dominated Sorting Genetic Algorithm II (NSGA-II) is employed to identify an optimal subset of EEG channels.

Link: doi.org/10.1007/978-3-031-78341-8_7

EEG based Brain Computer Interface: Development of mathematical and computational approaches for analysis of EEG-based brain-computer interface.

Explainable AI (XAI): Using XAI to spot the important areas of brain.

Graph Neural Networks (GNN): Exploring applications of GNN in understanding the structure of brain.

Adversarial Attacks: Creating attack and defense strategies to build a secure interface for AI.

Research Interests

Brain-Computer Interfaces, EEG Signal Processing, Explainable AI, and Graph Neural Networks, Adversarial Attacks, Artificial Intelligence.

Best Poster Award
Got the Best Poster Award for my thesis in Doctoral Colloquium at International Conference on Pattern Recognition and Machine Learning held at IIT Delhi (Dec, 2025).

Winner Hackathon
Got 1st Place in Hackathon that was conducted globally by NTX. The task was to come up with novel application based on EEG data. (Dec, 2023)

Winner Data Challenge
Got 1st Place in the data challenge that was conducted globally by NTX. The task was to identify sleep stages based on EEG data. (Jan 2024)

Aashish Khilnani

Assistant Professor

Ph.D. (Banaras Hindu University)

School of Computer Science and Engineering

1