Dorde Popovic
Scholar

Dorde Popovic

Google Scholar ID: mBtJH9sAAAAJ
Qatar Computing Research Institute, HBKU
Neural Trojan BackdoorsAdversarial LearningRobust Machine LearningFederated LearningEthics of Artificial Intelligence
Citations & Impact
All-time
Citations
26
 
H-index
2
 
i10-index
1
 
Publications
5
 
Co-authors
8
list available
Resume (English only)
Academic Achievements
  • Andrew Carnegie Society Scholar Award (CMU 2022); Outstanding Academic Achievement Award (CMUQ 2022); Allen Newell Award for Excellence in Undergraduate Research (Runner-up, CMU 2022); Best Technique and Data-Efficiency Award in the Trojan Detection Challenge (Runner-up, NeurIPS 2022); Best Project Award at the Meeting of the Minds Research Symposium (2nd and 3rd place, CMUQ 2022); Qatar Campus Scholar Award (CMUQ 2022). Publications: 'DeBackdoor: A Deductive Framework for Detecting Backdoor Attacks on Deep Models with Limited Data' (USENIX Security '25); 'FedTeams: Towards Trust-Based and Resource-Aware Federated Learning' (CloudCom 2022). Reviewer for IEEE Transaction on Dependable and Secure Computing (TDSC 2023) and ACM International World Wide Web Conference (WWW 2023).
Research Experience
  • Research Engineer in the Cyber Security Research Group at Qatar Computing Research Institute. Projects include: aiXamine (LLM Safety & Security, Simplified), Fanar (An Arabic-Centric Multimodal Generative AI Platform), DeBackdoor (A Deductive Framework for Detecting Backdoor Attacks with Limited Data), FedTeams (Towards Trust-Based and Resource-Aware Federated Learning), CASPRE (A CRISPR-Cas12a–Powered Handheld Kit for Rapid Genetic Carrier Testing).
Education
  • BSc in Computer Science from Carnegie Mellon University, 2018-2022.
Background
  • Research interests include backdoor attacks and defenses in deep learning, robust machine learning, security of LLMs, federated learning, and the ethics of AI.