I am a fifth-year Ph.D. candidate in the School of Electrical and Computer Engineering at Purdue University, advised by Prof. Saurabh Bagchi in the Dependable Computing Systems Laboratory (DCSL) with an expected graduation of May 2026. Previously, I received my M.S. in Computer Science and B.S. in Computer Engineering from Northwestern University in 2021.

My research focuses on machine learning with an emphasis on building scalable, reliable, and robust systems. As a machine learning security researcher, I am particularly passionate when learning about new systems and improving the safety (reliability of the models and the ethical uses) of real world deep learning applications.

My work has been published in top computer vision, security, and systems conferences (CVPR, IEEE S&P, AsiaCCS, DSN-S), ranging from data privacy and model reliability under data heterogeneity (key problems in real-world deployments of large-scale settings in federated learning) to model robustness (reliability, adversarial machine learning, and efficient robust transfer learning). One of my recent works also explores using foundation models to improve weakly supervised point cloud semantic segmentation.

I have also worked in autonomous driving / ADAS perception at Woven by Toyota, where I trained and evaluated models for occupancy prediction and scene flow.

Here is a summary of my research.

Selected publications

For a full list of publications (along with YouTube videos and paper links), see the Publications page. My Google Scholar can also be found here.

LOKI: Large-scale Data Reconstruction Attack against Federated Learning through Model Manipulation
Joshua C. Zhao, Atul Sharma, Ahmed Roushdy Elkord, Yahya H. Ezzeldin, Salman Avestimehr, Saurabh Bagchi
The 45th IEEE Symposium on Security and Privacy (S&P 2024)

Leak and Learn: An Attacker’s Cookbook to Train Using Leaked Data from Federated Learning
Joshua C. Zhao, Ahaan Dabholkar, Atul Sharma, Saurabh Bagchi
The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2024)

The Resource Problem of Using Linear Layer Leakage Attack in Federated Learning
Joshua C. Zhao, Ahmed Roushdy Elkordy, Atul Sharma, Yahya H. Ezzeldin, Salman Avestimehr, Saurabh Bagchi
The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2023)

The Federation Strikes Back: A Survey of Federated Learning Privacy Attacks, Defenses, Applications, and Policy Landscape
Joshua C. Zhao, S. Bagchi, S. Avestimehr, K. Chan, S. Chaterji, D. Dimitriadis, J. Li, N. Li, A. Nourian, H. Roth
ACM Computing Surveys (CSUR 2025)

Awards, Services, and Teaching:

• Bilsland Dissertation Fellowship
• Purdue Andrews fellowship
• DCSL Best Fresher Award, Group Champion Award
• Eta Kappa Nu, Beta Tau Chapter (Electrical Engineering Honor Society)
• Tau Beta Pi (Engineering Honor Society)
• Northwestern University SURG

Guest lecturer at Purdue University for ECE 60872 (Fault-Tolerant Computer System Design). I gave the two following lectures on ML security:

• How reliable is your model? (adversarial machine learning)
• Distributed machine learning, a secure and private alternative?

Reviewer: ICCV, CVPR, CVPR FedVision, NeurIPS, AAAI