Publications

In Drafts / In Submission


(*) Equal Contribution

[X.4] Breaking Free: How to Hack Safety Guardrails in Black-Box Diffusion Models!

Authors: Shashank Kotyan*, PoYuan Mao*, and Danilo Vasconcellos VargasIn Submission( 📝 Blog Post)(🌐 Project Webpage)
(📇 Link to pre-print article)

[X.3] The Challenges of Image Generation Models in Generating Multi-Component Images

Authors: Tham Yik Foong, Shashank Kotyan, PoYuan Mao, and Danilo Vasconcellos VargasIn Submission (📇 Link to pre-print article)

[X.2] Synthetic Shifts to Initial Seed Vector Exposes the Brittle Nature of Latent-Based Diffusion Models

Authors: PoYuan Mao*, Shashank Kotyan*, Tham Yik Foong, and Danilo Vasconcellos VargasIn Submission (📇 Link to pre-print article)

[X.1] k* Distribution: Evaluating the Latent Space of Deep Neural Networks using Local Neighborhood Analysis

Authors: Shashank Kotyan*, Ueda Tatsuya*, Danilo Vasconcellos VargasIn Revision in IEEE Transactions on Neural Networks and Learning Systems(🌐 Project Webpage)
(👨‍💻Link to Code) (📇 Link to pre-print article)

2024

[U.3] Towards Improving Robustness Against Common Corruptions in Object Detectors Using Adversarial Contrastive Learning

Authors: Shashank Kotyan, and Danilo Vasconcellos VargasIn Drafts(📇 Link to pre-print article)

[U.2] Towards Improving Robustness Against Common Corruptions using Mixture of Class Specific Experts

Authors: Shashank Kotyan, and Danilo Vasconcellos VargasIn Drafts(📇 Link to pre-print article)

[J.4] Improving Robustness for Vision Transformer with a Simple Dynamic Scanning Augmentation

Authors:  Shashank Kotyan, and Danilo Vasconcellos VargasPublished in Neurocomputing
                                                                                                                                                                                                                                              (📇 Link to pre-print article) (📢 Link to Published Article)

2023

[U.1] A reading survey on adversarial machine learning: Adversarial attacks and their understanding

Authors: Shashank Kotyan(📇 Link to pre-print article)

[PT] Towards Developing Robust Neural Networks By Understanding Adversarial Attacks and Defences

Author: Shashank Kotyan (👨‍🎓Link to Thesis) 

2022

[J.3] Transferability of features for neural networks links to adversarial attacks and defences

Authors:  Shashank Kotyan, Moe Matsuki, and Danilo Vasconcellos VargasPublished in PLOS ONE(📰 University Press Release JP PDF) (📰 University Press Release JP) (📰 University Press Release EN PDF) (📰 University Press Release EN)(👨‍💻Link to Code) (📇 Link to pre-print article) (📢 Link to Published Article)

[J.2] Adversarial Robustness Assessment: Why Both L0 And L∞ Attacks Are Necessary

Authors: Shashank Kotyan , Danilo Vasconcellos VargasPublished in PLOS ONE(👨‍💻Link to Code) (📇 Link to pre-print article) (📢 Link to Published Article) 

2021

[W.2] Deep Neural Network Loses Attention To Adversarial Images

Authors:  Shashank Kotyan, and Danilo Vasconcellos VargasPublished in 2021 at Workshop on Artificial Intelligence Safety (AISafety 2021) at the 30th International Joint Conference on Artificial Intelligence (IJCAI 2021)(🗣 Link to Presentation) (🗒 Link to Slides) (📇 Link to pre-print article) (📢 Link to Published Article)

2020

[W.1] Evolving Robust Neural Architectures To Defend From Adversarial Attack

Authors:  Shashank Kotyan, and Danilo Vasconcellos VargasPublished in 2020 at Workshop on Artificial Intelligence Safety (AISafety 2020) AT the 29th International Joint Conference on Artificial Intelligence and the 17th Pacific Rim International Conference on Artificial Intelligence (IJCAI-PRICAI 2020) (🗣 Link to Presentation) (🗒 Link to Slides) (👨‍💻Link to Code) (📇 Link to pre-print article) (📢 Link to Published Article)

[C.5] Is Neural Architecture Search A Way Forward To Develop Robust Neural Networks?

Authors: Shashank Kotyan, and Danilo Vasconcellos VargasPublished in 2020 at the 34th Annual Conference of the Japanese Society for Artificial Intelligence (JSAI 2020)(👨‍💻Link to Code) (📢 Link to Published Article)

[C.4] Towards Evolving Robust Neural Architectures To Defend From Adversarial Attacks

Authors: Shashank Kotyan, and Danilo Vasconcellos VargasPublished in 2020 at the 21st Genetic and Evolutionary Computation Conference Companion (GECCO 2020).(👨‍💻Link to Code) (📢 Link to Published Article)

2019

[UT] Understanding Vulnerabilities Of Deep Neural Networks To Evolve Robust Neural Architectures

Author: Shashank Kotyan  (👨‍🎓Link to Thesis)  

[C.3] Self Training Autonomous Driving Agent 

Authors:  Shashank Kotyan, Danilo Vasconcellos Vargas,  and Venkanna Udutalapally Published in 2019 at the 58th Annual Conference of the Society of Instrument and Control Engineers of Japan (SICE 2019)(📇 Link to pre-print article) (📢 Link to Published Article)  

2018

[C.2] Drishtikon: An Advanced Navigation System For Visually Impaired People

Authors: Shashank Kotyan, Venkanna Udutalapally, Nishant Kumar, and Pankaj Kumar SahuPublished in 2018 at the 2nd Conference on Information and Communication Technology (CICT 2018)  (📇 Link to pre-print article) (📢 Link to Published Article)

[C.1] HAUAR: Home Automation Using Action Recognition

Authors: Shashank Kotyan, Nishant Kumar, Pankaj Kumar Sahu,  and Venkanna UdutalapallyPublished in 2018 at the 2nd Conference on Information and Communication Technology (CICT 2018) (📇 Link to pre-print article) (📢 Link to Published Article)

[J.1] Analysis And Comparison Of Different Fuzzy Inference Systems Used In Decision Making For Secondary Users In Cognitive Radio Network

Authors: Shrivishal Tripathi, Ashish Upadhyay, Shashank Kotyan,  and Sandeep YadavPublished in Volume 104, Issue 3 of Wireless Personal Communication(📇 Link to pre-print article) (📢 Link to Published Article)

"The road to wisdom? -- Well, it's plain and simple to express: Err and err and err again but less and less and less."

-Piet Hein

Last Updated: June 2024