I completed my Ph.D. in Computer Science at UC San Diego jointly advised by Professor Julian McAuley and Professor Shlomo Dubnov. My Ph.D. thesis focused on synthesis and robust detection of AI generated media. I received my undergraduate (B. Tech) degree in Computer Science from Indian Institute of Technology (IIT), Roorkee in 2017.
Currently, I am a Research Scientist at NVIDIA, working on large-scale speech synthesis models. Additionally, I provide mentorship to the technical team at Blue Water, an organization dedicated to developing fuel optimization solutions for commercial shipping companies.

News

[May 2024] - Our paper SelfVC: Voice Conversion With Iterative Refinement using Self Transformations got accepted to ICML 2024!
[Apr 2024] - Our paper on watermarking LLM generated texts REMARK-LLM: A Robust and Efficient Watermarking Framework for Generative Large Language Models got accepted to Usenix Security 2024!
[Jan 2024] - FaceSigns got accepted to ACM Transactions on Multimedia Computing, Communications, and Applications! SemiFragile Neural Watermarking to guard against DeepFakes. Try out our demo
[Jun 2023] - Started a new position as a Research Scientist, Conversational AI at NVIDIA !
[May 2023] - I defended my PhD thesis! You can access my talk here
[Feb 2023] - Our paper ACE-VC: Adaptive and Controllable Voice Conversion using Explicitly Disentangled Self-supervised Speech Representations got accepted to ICASSP 2023. Checkout the voice conversion examples !
[Sep 2022] - Our paper FastStamp: Accelerating Neural Steganography and Digital Watermarking of Images on FPGAs got accepted to ICCAD 2022.
[Oct 2021] - Our work on Cross-modal Adversarial Reprogramming got accepted to WACV 2022!
[Sep 2021] - Our work on Expressive Voice Cloning got accepted to ACML 2021! Try cloning your own voice by recording just 10 utterances using our demo or listen to some audio examples.

[Jun 2021] - Started my summer internship with NVIDIA’s Text to Speech Synthesis team.
[Jan 2021] - Our paper WaveGuard: Understanding and mitigating audio adversarial examples got accepted to USENIX Security 2021.
[Nov 2020] - Our paper Adversarial Deepfakes: Evaluating Vulnerability of Deepfake Detectors to Adversarial Examples got accepted to WACV 2021.
[Sep 2020] - Completed summer research internship with Facebook AI Red Team. Worked on exploring vulnerabilities of DeepFake detectors.
[Aug 2019] - Our paper Adversarial Reprogramming of Text Classification Neural Networks got accepted at EMNLP 2019.
[Jun 2019] - Two papers accepted at INTERSPEECH 2019.
[Jun 2019] - Our paper FastWave: Accelerating Autoregressive Convolutional Neural Networks on FPGA got accepted at ICCAD 2019.

Publications

Improving Robustness of LLM-based Speech Synthesis by Learning Monotonic Alignment
Paarth Neekhara*, Shehzeen Hussain*, Subhankar Ghosh, Jason Li, Rafael Valle, Rohan Badlani, Boris Ginsburg
Interspeech, 2024
[ pdf, arxiv, audio examples ]

SelfVC: Voice Conversion With Iterative Refinement using Self Transformations
Paarth Neekhara*, Shehzeen Hussain*, Rafael Valle, Boris Ginsburg, Rishabh Ranjan, Shlomo Dubnov, Farinaz Koushanfar, Julian McAuley
ICML, 2024
[ pdf, arxiv, demo ]

REMARK-LLM: A Robust and Efficient Watermarking Framework for Generative Large Language Models
Ruisi Zhang, Shehzeen Samarah Hussain, Paarth Neekhara, Farinaz Koushanfar
Usenix Security, 2024
[ pdf, arxiv ]

FaceSigns: Semi-fragile watermarks for media authentication
Paarth Neekhara*, Shehzeen Hussain*, Xinqiao Zhang, Farinaz Koushanfar, Julian McAuley
ACM Transactions on Multimedia Computing, Communications, and Applications, 2024
[ pdf, arxiv, demo ]

ReFace: Real-time Adversarial Attacks on Face Recognition Systems
Shehzeen Hussain, Todd Huster, Chris Mesterharm, Paarth Neekhara, Kevin An, Malhar Jere, Harshvardhan Sikka, Farinaz Koushanfar
DSN 2023
[ pdf, arxiv ]

ACE-VC: Adaptive and Controllable Voice Conversion using Explicitly Disentangled Self-supervised Speech Representations
Shehzeen Hussain*, Paarth Neekhara*, Jocelyn Huang, Jason Li, Boris Ginsburg
ICASSP 2023
[ pdf, arxiv, audio examples ]

FastStamp: Accelerating Neural Steganography and Digital Watermarking of Images on FPGAs
Shehzeen Hussain*, Nojan Sheybani*, Paarth Neekhara*, Xinqiao Zhang, Javier Duarte, Farinaz Koushanfar
ICCAD 2022
[ pdf, arxiv ]

Cross-modal Adversarial Reprogramming
Paarth Neekhara*, Shehzeen Hussain*, Jinglong Du, Shlomo Dubnov, Farinaz Koshanfar, Julian McAuley
WACV 2022
[ pdf, arxiv ]

Expressive Neural Voice Cloning
Paarth Neekhara*, Shehzeen Hussain*, Shlomo Dubnov, Farinaz Koshanfar, Julian McAuley
ACML 2021
[ pdf, arxiv, audio examples, demo ]

Adversarial threats to deepfake detection: A practical perspective
Paarth Neekhara, Brian Dolhansky, Joanna Bitton, Cristian Ferrer
CVPR Media Forensics Workshop 2021
[ pdf, arxiv ]

WaveGuard: Understanding and mitigating audio adversarial examples
Shehzeen Hussain*, Paarth Neekhara*, Shlomo Dubnov, Julian McAuley, Farinaz Koshanfar
USENIX Security 2021
[ pdf, arxiv ]

Adversarial Deepfakes: Evaluating Vulnerability of Deepfake Detectors to Adversarial Examples
Shehzeen Hussain*, Paarth Neekhara*, Malhar Jere, Farinaz Koshanfar, Julian McAuley
WACV 2021
[ pdf, arxiv, website ]

Adversarial Reprogramming of Text Classification Neural Networks
Paarth Neekhara, Shehzeen Hussain, Shlomo Dubnov, Farinaz Koshanfar
EMNLP 2019
[ pdf, arxiv, code ]

Universal Adversarial Perturbations for Speech Recognition Systems
Paarth Neekhara*, Shehzeen Hussain*, Prakhar Pandey, Shlomo Dubnov, Julian McAuley, Farinaz Koushanfar
INTERSPEECH 2019
[ pdf, arxiv, audio examples ]

Expediting TTS Synthesis with Adversarial Vocoding
Paarth Neekhara*, Chris Donahue*, Miller Puckette, Shlomo Dubnov, Julian McAuley
INTERSPEECH 2019
[ pdf, arxiv, code, audio examples ]

FastWave: Accelerating Autoregressive Convolutional Neural Networks on FPGA
Shehzeen Hussain, Mojan Javaheripi, Paarth Neekhara, Ryan Kastner, Farinaz Koushanfar
ICCAD 2019
[ pdf, arxiv ]