Mohammadreza Davoodi Behbahani

AI Developer and Researcher

I am currently a non-profit researcher in the Data Science and Machine Learning Laboratory.
I hold an MSc in Artificial Intelligence from Iran University of Science and Technology, specializing in Image Processing and Computer Vision. Prior to that, I earned a Bachelor’s degree in Electrical Engineering with a focus on Communication from Isfahan University of Technology.
Passionate about designing and implementing smart systems, I enjoy working on innovative solutions that integrate AI and advanced technologies.

Download CV

IELTS Result: 7.0 Overall: {Listening:7.5, Reading:7.5, Writing: 6.0, Speaking: 7.0}

Age

Email

me[AT]mrdavoodi[DOT]ir

Phone

(+98) 937-029-0064

MrDavoodi

Scholar

Education

Masters in Computer Engineering (Artificial Intelligence)

September 2019 – February 2023

Thesis Title: One-shot Similarity on Image Recognition
Iran University of Science and Technology

Bachelor in Electrical Engineering

September 2015 – September 2019

Isfahan University of Technology

Publications

RealDrag: The First Dragging Benchmark with Real Target Image

Ahmad Zafarani, Zahra Dehghanian, Mohammadreza Davoodi, Mohsen Shadroo, MohammadAmin
Fazli, Hamid R. Rabiee

arXiv preprint, arXiv:2512.12287, 2025. (Submitted to CVPR 2026)

Research Projects

Multimodal Analysis of Movie Data (Ongoing Project, Intended Submission: KDD 2026)

Performing multimodal analysis on movie data from multiple sources, including preprocessing, feature extraction, and statistical analysis across text, image, and video modalities. The goal is to model and predict content ratings using multimodal signals.

Fake News Detection Using Multimodal Data (Ongoing Project, Intended Submission: ACL 2026)

Investigating multimodal fake news detection using text and video. Currently benchmarking prior works and evaluating models such as BERT, CLIP, and multi-agent frameworks for cross-modal reasoning.

Deepfake Detection in Videos

Designed a deepfake detection pipeline combining EﬃcientNet, Vision Transformer, GenConvViT,
AltFreezing, and UCF methods. Built a custom dataset of manipulated videos, achieving 98% accuracy with UCF, and explored ensemble learning and Qwen2-VL for future extensions.

Logo Detection Using YOLOv5

Created a dataset of about 100 logo categories (brands and TV channels) and ﬁne-tuned YOLOv5 for
multi-class detection. Achieved over 80% accuracy and analyzed challenges such as dataset imbalance and logo variability.

Retrieval-Augmented Generation (RAG) Service with LLMs

Developed a document-based question answering system using Qwen-2B, integrating a retrieval pipeline to improve factual accuracy in responses.

One-shot Similarity on Image Recognition (MSc Thesis)

Fine-tuned YOLOv5 (ImageNet weights) for 16-class image recognition and implemented a one-shot
similarity method for four additional classes. Studied the eﬀect of varying ﬁne-tuned layers on performance.

Skills

Python

Pytorch, Tensorflow, Opencv, Numpy, Scipy, Scikit learn, Matplotlib, Pandas

C and C++

Matlab

Linux

Git

Docker/Kubernetes

Research Interests

These are the areas I’m interested to work.

Vision-Language Models (VLMs) and Multimodal AI

Deepfake Detection and Multimedia Forensics

Computer Vision and Image Processing

Natural Language Processing (NLP) and Text Understanding

Large Language Models (LLMs) and Generative AI

Applying Deep Learning on Bioinfrmatics Data