Sayanton Dibbo | The University of Alabama

Research Overview

My research focuses on the intersection of AI/ML (Generative AI, Computer Vision, LLM, Audio Modeling), and Trustworthy AI. The TAI Lab at UA investigates how to make foundational models (LLMs and Multimodal systems) more robust against adversarial and privacy attacks.

Current Focus: We are currently analyzing vulnerabilities of foundational models and developing novel defense techniques to enhance the robustness of foundational models under different privacy and adversarial attacks.
View Project Descriptions →

Selected Publications

Are Neuro-Inspired Multi-Modal Vision-Language Models Resilient to Membership Inference Privacy Leakage? Paper teaser

David Amebley, Sayanton Dibbo

Pre-print (*Under Review)

PDF Code BibTeX

Improving Robustness to Model Inversion Attacks via Sparse Coding Architectures Paper teaser

Sayanton Dibbo, Adam Breuer, Juston Moore, Michael Teti

The 18th European Conference on Computer Vision (ECCV), 2024

PDF Code BibTeX

Full List

Publications

Pre-print

Are Neuro-Inspired Multi-Modal Vision-Language Models Resilient to Membership Inference Privacy Leakage?

David Amebley, Sayanton Dibbo,

USENIX WOOT Conference(USENIX WOOT), 2026.

PDF arXiV BibTex Under Review

ECCV

Improving Robustness to Model Inversion Attacks via Sparse Coding Architectures

Sayanton Dibbo, Adam Breuer, Juston Moore, Michael Teti

The 18th European Conference on Computer Vision (ECCV), 2024.

PDF arXiV Project Page BibTex Published

ICASSPW

LCANets++: Robust Audio Classification using Multi-layer Neural Networks with Lateral Competition

Sayanton Dibbo, Juston Moore, Garrett Kenyon, Michael Teti

IEEE International Conference on Acoustics, Speech and Signal Processing Workshops (ICASSPW), 2024.

PDF arXiV Project Page BibTex Published

CSF

SoK: Model Inversion Attack Landscape: Taxonomy, Challenges, and Future Roadmap

Sayanton Dibbo

IEEE 36th Computer Security Foundations Symposium (CSF), 2023.

PDF BibTex Slides Published

AdvML

How Can Neuroscience Help Us Build More Robust Deep Neural Networks?

Sayanton Dibbo, Siddharth Mansingh, Jocelyn Rego, Garrett Kenyon, Juston Moore, Michael Teti

ICML Workshop on Adversarial Machine Learning (AdvML), 2023.

PDF BibTex Video Published

TDSC

Bag of on-phone ANNs to secure IoT objects using wearable and smartphone biometrics

Sudip Vhaduri, William Cheung, Sayanton Dibbo

IEEE Transactions on Dependable and Secure Computing (TDSC), 2023.

PDF BibTex Published

SaTML

Model Inversion Attack with Least Information and an In-depth Analysis of its Disparate Vulnerability

Sayanton Dibbo, Dae Lim Chung, Shagufta Mehnaz

IEEE Conference on Secure and Trustworthy Machine Learning (SaTML), 2023.

PDF BibTex Video Published

USENIX

Are Your Sensitive Attributes Private? Novel Model Inversion Attribute Inference Attacks on Classification Models

Shagufta Mehnaz, Sayanton Dibbo, Ehsanul Kabir, Ninghui Li, Elisa Bertino

31st USENIX Security Symposium (USENIX Security), 2022.

PDF arXiV BibTex Video Published

SaSeIoT

On-phone CNN model-based implicit authentication to secure IoT wearables

Sayanton Dibbo, William Cheung, Sudip Vhaduri

The 5th International Conference on Safety and Security with IoT (SaSeIoT), 2022.

PDF arXiV BibTex Best Presentation