NPerseus
Networked research · DTU CSE
Secure AI · Privacy · Digital Forensics

Secure intelligence. Verifiable by design.

A research group in the Department of Computer Science & Engineering at Delhi Technological University working at the intersection of deep learning, information security and multimedia forensics — building models you can trust and methods that keep data safe.

Adversarial MLReversible Data HidingImage SteganographyDeepfake ForensicsAI Privacy & SecurityFederated LearningTrustworthy AIBackdoor Defense
The NPerseus model

Research is stronger when every result has a trail.

  1. Connect

    People, projects and shared research infrastructure.

  2. Verify

    Evidence, experiments and contributions with provenance.

  3. Advance

    Reusable knowledge that compounds across the lab.

129Reviewed publications56 journal · 62 conference
19Patent records15 granted · 4 published
15Active people3 faculty · 12 researchers
6Ph.D. scholarsAcross the research team
4Research directionsSecurity, privacy & forensics
3AlumniPart of the lab network

23 listed works in 2026 · Publications are deduplicated across faculty and members. Patent statuses follow source records. Browse the evidence →

Faculty

Three perspectives. A shared research programme.

Computer Science & Engineering, Delhi Technological University.

Meet the team →
Dr. Rajeev Kumar

Dr. Rajeev Kumar

Assistant Professor

Leads the lab's work on security and privacy of deep neural networks, data hiding and multimedia forensics. Ph.D. (University of Delhi, 2017) with a post-doctorate at Kyungil University, South Korea and an NRF Korea Brain Pool Plus Fellowship.

Privacy & Security of Deep Neural NetworksReversible Data Hiding & WatermarkingImage & Text Steganography
Dr. Anurag Goel

Dr. Anurag Goel

Assistant Professor

Works on machine learning, deep learning and deep clustering, graph algorithms, image processing, AI for social good and computational creativity. He earned his Ph.D. in CSE from IIIT-Delhi in 2023 with a thesis on Deep Clustering and received IIIT-Delhi’s M.Tech. Gold Medal in 2017.

Machine LearningDeep LearningDeep Clustering
Ms. Anukriti Kaushal

Ms. Anukriti Kaushal

Assistant Professor

Researches manipulated-image detection, deepfake generation and detection, and digital image integrity. M.Tech (DTU) and a Ph.D. in progress at DTU on intelligent image-forensics strategies.

Image ForensicsDeepfake DetectionDeep Learning
Questions, methods & evidence

Inside the research.

Projects & technology →

Faculty project portfolio

All source-listed projects →
Research output

Recent publications.

Latest year first, with authors, publication type and original sources.

All 129 records →
Research focus · security × data hiding

Reversible adversarial perturbations

Studying perturbations that can be removed to recover the original image, connecting adversarial robustness with reversible data hiding.

A Survey on Reversible Adversarial Perturbations

Nikhil Gupta, Ankur, Rajeev Kumar · 2026

Conference paper

Read the source-linked survey ↗
The people

Researchers driving the work.

Full team →

6 of 12 researchersExplore at your pace

Ayush Kumar

B.Tech Student Researcher

Android Security, Vulnerability Research and Secure Systems

Platform engineering

Built and maintains the NPerseus website and research tools.

NPerseus

Ajit Kumar Yadav

Full-time Ph.D. Candidate

Reversible Data Hiding in Encrypted Images, Image Security, Backdoor Detection in Neural Networks

Latest listed publication

Adaptive Weighted Predictor for Checkerboard based Reversible Data Hiding in Encrypted Images

2026 · Conference paper

Mahaveer Prasad

M.Tech Student

Adversarial Machine Learning, Backdoor and Data-Poisoning Defense, Network Intrusion Detection, Android Malware Analysis

Latest listed publication

A Unified Multi-Signal Indicator Framework for Backdoor Detection and Label Correction in Deep Neural Networks

2026 · Conference paper

Nikhil Gupta

Full-time Ph.D. Candidate

Enabling Technologies for Trustworthy Artificial Intelligence, Adversarial Machine Learning, AI Security and Robustness

Latest listed publication

A Survey on Reversible Adversarial Perturbations

2026 · Conference paper

Beyond the paper

Patents, recognition & lab updates.

All news & activity →
Work with NPerseus

Turn a hard trust problem into research we can test.

Collaborate on secure AI, sponsor a focused research project, or join the researchers building verifiable intelligent systems at DTU.