Ph.D. Candidate (CSE), IIT Jammu • Visiting Scholar (NTNU) • Reviewer @ CVPR, WACV, ICPR • Publications at IJCNN 2025, WACV 2024/2025, IEEE SMC, BigData, and more.
I am a researcher focused on privacy-preserving biometrics and image processing. My work explores synthetic biometric generation, cancelable templates, and inpainting/outpainting models that contribute to secure authentication systems and robust visual understanding.
LinkedIn update about my presentation at the 22nd International Conference on Distributed Computing and Intelligent Technology (ICDCIT 2026), held in Bhubaneswar from 16–19 January 2026.
Read post →
Quick update from my LinkedIn regarding the IJCNN 2025 INNS Doctoral Consortium. Click to read the post.
Read post →
Quick update from my LinkedIn regarding IEEE SMC 2024. Click to read the post.
Read post →
Quick update from my LinkedIn about attending and presenting my research. Click to read the post.
Read post →
Short note from my LinkedIn about my time at the Norwegian University of Science and Technology. Click to read the post.
Read post →
Short update from my LinkedIn about participating in a prestigious program. Click to read the post.
Read post →
Short update from my LinkedIn about IJCB 2025. Click to read the post.
Read post →
Frequency-Split Conditioning for Diffusion-Based Contactless Fingerprint Rendering.
our paper was accepted and presented at the 21st International Conference on Computer Vision Theory and Applications (VISAPP 2026).
Developed proxy fingerprints combined with user-specific keys for revocability and cancelability.
Custom generative model architecture to generate biometric data with controllable variation.
Convolutional network for deformed fingerprint restoration via image inpainting.
Key-based generative approach to create revocable, non-invertible proxy fingerprint templates that preserve matching utility while protecting the original biometric.
Pipeline combining class-transformation and user-specific key projection to generate realistic, revocable fingerprint templates; evaluated with MCC matcher.
A multi-stage framework that starts by first providing lower-resolution feature representations from corrupted frames using a context encoder.
For collaborations, research discussions, or speaking engagements.