Mohammad Reza Taesiri

I am currently a second third-year Ph.D. student at the University of Alberta in Edmonton, focusing on the intersection of computer vision, explainable machine learning, and video games.


profile photo

Research

I am broadly interested in computer vision and machine learning, with a focus on explainable machine learning. I am also interested in applying Foundation models to real-world problems, such as video game testing and quality control.

Recent Highlights


Zoom  is what you need: An empirical study of the power of zoom and spatial biases in image classification GlitchBench: Can large multimodal models detect video game glitches?
Mohammad Reza Taesiri, Anh Nguyen Cor-Paul Bezemer,
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024

[Website] [Code] [Dataset]

Zoom  is what you need: An empirical study of the power of zoom and spatial biases in image classification ImageNet-Hard: The Hardest Images Remaining from a Study of the Power of Zoom and Spatial Biases in Image Classification aka. "Zoom is what you need: An empirical study of the power of zoom and spatial biases in image classification"
Mohammad Reza Taesiri, Giang Nguyen, Sarra Habchi, Cor-Paul Bezemer, Anh Nguyen
Conference on Neural Information Processing Systems (NeurIPS), 2023

[Website] [Code] [Dataset] [Dataset-4K]

Visual correspondence-based explanations Visual correspondence-based explanations improve AI robustness and human-AI team accuracy
Mohammad Reza Taesiri *, Giang Nguyen*, Anh Nguyen (* denotes equal contribution)
Conference on Neural Information Processing Systems (NeurIPS), 2022

[Website] [Live Demo] [Code] [Video]

CLIP Meets GamePhysics CLIP meets GamePhysics: Towards bug identification in gameplay videos using zero-shot transfer learning
Mohammad Reza Taesiri, Finlay Macklon, Cor-Paul Bezemer
The Mining Software Repositories (MSR) conference, 2022

[Website] [Code] [Live Demo] [Dataset]

Large Language Models are Pretty Good Zero-Shot Video Game Bug Detectors
Mohammad Reza Taesiri, Finlay Macklon, Yihe Wang, Hengshuo Shen, Cor-Paul Bezemer
ArXiv Preprint, 2022

[Website] [Code] [Dataset]

Website design by Jon Barron.