Education
| Degree | University | GPA | Thesis |
|---|---|---|---|
| MSc in Computer Engineering (Artificial Intelligence & Robotics) | Sharif University of Technology (SUT) | 18.40/20 (4.00/4.00) | Blind Image Super-Resolution using Deep Generative Neural Network Architectures |
| BSc in Computer Engineering | Amirkabir University of Technology (AUT) | 19.09/20 (3.96/4.00) | Graph-based Convolutional Multivariate Time Series Forecasting Approach for Urban Traffic Forecasting |
Research Interests
- Deep Generative Models (VAEs, GANs, Diffusion Models)
- Image and Video Segmentation
- Image/Video Restoration
- 3D Vision Reconstruction (e.g., NeRF, Gaussian Splatting)
- Computer Vision
- Knowledge Distillation
- Natural Language Processing
- Graph Neural Networks
Publications and Ongoing Work
KeyVIS: Improving Weakly-supervised Video Instance Segmentation Using Keypoints Consistency
Submitted to Iranian Machine Vision and Image Processing Conference (MVIP), October 2025ConsistVIS: Weakly-Supervised Video Instance Segmentation via Embedding Vector Consistency
Submitted to IEEE Transactions on Multimedia Journal, September 2025A Comprehensive Survey on Knowledge Distillation
Published in Transactions on Machine Learning Research (TMLR), September 2025 (DOI) (arXiv)CLBSR: A Deep Curriculum Learning-based Blind Image Super-Resolution Network using Geometrical Prior
Published in Image and Vision Computing Journal, February 2025 (DOI)GSCINet: Graph-based Convolutional Multivariate Time Series Forecasting Approach for Urban Traffic Forecasting
For a full list of my publications, visit my Google Scholar.
Teaching Assistance
| Date | Course | Supervisor | University |
|---|---|---|---|
| Fall 2025 | Advanced 3D Computer Vision | Prof. Kasaei | SUT |
| Spring 2025 | Digital Image Processing | Prof. Kasaei | SUT |
| Fall 2024 | Deep Learning | Prof. Beigy | SUT |
| Fall 2024 | Advanced 3D Computer Vision | Prof. Kasaei | SUT |
| Spring 2024 | Fundamental of 3D Computer Vision | Dr. Naderi | SUT |
| Spring 2023 | Data Mining | Prof. Nazerfard | AUT |
| Spring 2023 | Applied Linear Algebra | Prof. AmirMazlaghani | AUT |
| Spring 2022 | Signals and Systems | Dr. TermehChi | AUT |
| Fall 2022 | Applied Linear Algebra | Prof. Nazerfard | AUT |
| Fall 2022 | Data Structures and Algorithms | Prof. Shirali Shahreza | AUT |
Professional Experience
| Position | Organization | Duration | Description |
|---|---|---|---|
| Data Scientist | Bale Messenger | May 2023 - Feb. 2024 | Contributed to projects including channel classifier, recommender systems, and intelligent advertisement systems. |
| Machine Learning Engineer | Asr Gooyesh Pardaz | Jul. 2022 - Sep. 2022 | Developed audiovisual speech recognition for Persian language and created Persian audiovisual datasets. |
Projects
| Project | Description | Link |
|---|---|---|
| Deep Learning Homework | Includes dimensionality reduction, autoencoders, graph embeddings, and reinforcement learning tasks using methods like PCA, t-SNE, VAE | View on GitHub |
| Deep Generative Models Homework | Implementations of VAEs, GANs, Diffusion Models (DDPMs), and Energy-Based Models (EBMs) on datasets like MNIST and CIFAR-10 | View on GitHub |
| Digital Image Processing Homework | Solutions to assignments covering filtering, transformations, compression, and segmentation | View on GitHub |
| Panorama | Implementation of image stitching from scratch using the OpenCV library to create panoramic images.Implementation of image stitching from scratch using the OpenCV library to create panoramic images. | View on GitHub |
| URL-Shortener-as-a-Service | Developed a simple URL shortener service configured to run in a Kubernetes cluster. | View on GitHub |
| Information Retrieval Notebooks | Completed assignments for the Information Retrieval course at Amirkabir University of Technology, including a final project using ElasticSearch to implement Boolean Queries, Similarity Modulation, and Spell Correction. | View on GitHub |
| Monitoring System | Developed a monitoring system for the Computer Networks course to track system states and store them in Prometheus. | View on GitHub |
For a full list of my projects, visit my GitHub profile.
Skills and Expertise
- Programming Languages: Python, Java, C/C++, MATLAB
- Frameworks & Libraries: PyTorch, PyTorch Geometric, TensorFlow, Hugging Face, JAX, OpenCV
- Tools: Git, Linux, Bash, CUDA, FFmpeg, LangChain, AWS EC2, $\LaTeX$
- Data Science: Numpy, Pandas, Scikit-learn, Matplotlib, Gradio
Languages
- English
- TOEFL iBT: 101
Reading Listening Speaking Writing 28 27 23 23
- TOEFL iBT: 101
- Persian (Native)
Contact Information
- Email: amir.m.babaei.academic@gmail.com
- LinkedIn: Amir Mohammad Babaei
- Google Scholar: Amir Mohammad Babaei
I am always open to discussing potential collaborative research. Please feel free to reach out!