Scholar
Bardia Safaei
Google Scholar ID: Q62Ur2QAAAAJ
Johns Hopkins University
Citations & Impact
Citations
96
H-index
5
i10-index
5
Publications
8
Co-authors
11
list available
Publications
Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Academic Achievements
- Papers published:
- - One paper accepted at MICCAI '25
- - One paper accepted at ICIP '25
- - Two papers accepted at CVPR '25 as Highlights: Filter Images First, Generate Instructions Later: Pre-Instruction Data Selection for Visual Instruction Tuning, Towards Zero-Shot Anomaly Detection and Reasoning with MLLMs
- - Our paper Active Learning for Vision-Language Models has been accepted at WACV '25, Tucson, Arizona, USA
- - Our paper Gradient-Regularized Out-of-Distribution Detection has been accepted at ECCV '24, Milano, Italy
- - Our paper Entropic Open-set Active Learning has been accepted at AAAI '24, Vancouver, Canada
- - Our paper Open-set Automatic Target Recognition has been accepted at ICASSP '23, Rhodes Island, Greece
Research Experience
- Summer 2024 internship at Honda Research Institute focusing on data-efficient instruction tuning for multimodal LLMs (Mentor: Shao-Yuan Lo)
Education
- PhD in Electrical and Computer Engineering, Johns Hopkins University, Advisor: Prof. Vishal M. Patel (2021-Present)
- BSc in Electrical Engineering, Communication Systems, Sharif University of Technology (2016-2021)
Background
- Research interests include computer vision and natural language processing, with a specific focus on multimodal LLMs (reasoning, instruction tuning, hallucination), continual/active learning, large-scale data pruning, efficient training (prompting, token pruning, MoE), and out-of-distribution robustness.
Miscellany
- CV can be found here
