Mojtaba Nafez

I am a Master's student at Sharif University of Technology in Computer Engineering, supervised by Prof. MohammadHossein Rohban in the RIML Lab.

In recent years, I have been conducting research in the RIML Lab under the supervision of Prof. MohammadHossein Rohban, closely collaborating with Prof. Mohammad Sabokrou.

Previously, I received my Bachelor of Science in Computer Engineering from Iran University of Science and Technology. During the final year of my Bachelor's, I was a research assistant at CVLab IUST, working under the supervision of Prof. Mohammad Reza Mohammadi.

Email  /  Google Scholar  /  Github  /  LinkedIn

profile photo

Research

I have extensive research experience in trustworthiness and reliability machine learning models and have a strong academic interest in developing robust machine learning algorithms, especially within the domains of computer vision and natural language processing (NLP).

Publications & Preprints
PatchGuard: Adversarially Robust Anomaly Detection and Localization through Vision Transformers and Pseudo Anomalies
Mojtaba Nafez, Amirhossein Koochakian, Arad Maleki, MohammadHossein Rohban
CVPR, 2025
will be updated!

PatchGuard is an adversarially robust Anomaly Detection (AD) and Localization (AL) method using pseudo anomalies in a ViT-based framework. It leverages Foreground-Aware Pseudo-Anomalies and a novel loss function for improved robustness, achieving significant performance gains on industrial and medical datasets under adversarial settings.

Adversarially Robust Anomaly Detection through Spurious Negative Pair Mitigation
Hossein Mirzaei*, Mojtaba Nafez*, Jafar Habibi, Mohammad Sabokrou, MohammadHossein Rohban
ICLR, 2025
openreview

Anomaly Detection (AD) methods are vulnerable to adversarial attacks due to relying on unlabeled normal samples. The authors address this by creating a pseudo-anomaly group and using adversarial training with contrastive loss, mitigating spurious negative pairs through opposite pairs to improve robustness.

Universal Novelty Detection Through Adaptive Contrastive Learning
Hossein Mirzaei, Mojtaba Nafez, Mohammad Jafari, Mohammad Bagher Soltani, Mohammad AzizMalayeri, Jafar Habibi, Mohammad Sabokrou, MohammadHossein Rohban
CVPR, 2024
arXiv / code

Addressing a critical practical challenge within the domain of image-based anomaly detection, our research confronts the absence of a universally applicable and adaptable methodology that can be tailored to diverse datasets characterized by distinct inductive biases.

Scanning Trojaned Models Using Out-of-Distribution Samples
Hossein Mirzaei, Ali Ansari, Bahar Nia , Mojtaba Nafez, Moein Madadi, Sepehr Rezaee, Zeinab Taghavi, Arad Maleki, Kian Shamsaie, Hajialilue, Jafar Habibi, Mohammad Sabokrou, Mohammad Hossein Rohban
NeurIPS, 2024
arXiv / code

In this research, our problem was to identify whether a given model was backdoored or not. The study finds that backdoored models exhibit jagged decision boundaries around out-of-distribution (OOD) samples, leading to reduced robustness.

Honors & Awards
    Academic Ranking: 3rd among 105 students in Bachelor’s degree at Iran University of Science and Technology (IUST)

Design and source code from Jon Barron's website.