Sina Eskandari

Sina Eskandari

Computer Engineering Student at IUST, Interested in Deep Learning, Computer Vision and NLP, Ambitious Learner and Adventurer

Sina Eskandari

Hello, I’m Sina. I graduated from Iran University of Science and Technology(IUST) with Bachelor of Science in Computer Engineering. Currently, I work as a research assistant at Computer Vision Lab of IUST. My research interests are Deep Learning, Computer Vision, Natural Language Processing and Machine Learning. My journey involves unraveling the possibilities at the intersection of technology and artificial intelligence, contributing to the ever-evolving landscape with a solid foundation from my studies.

Education

  • Bachelor of Science in Computer Engineering, Iran University of Science and Technology, 2018 - 2023, Overall GPA: 3.37/4, Last Two Year GPA: 3.93/4

Research Interests

  • Computer Vision
  • Deep Learning
  • Natural Language Processing
  • Machine Learning

Research Experience

  • Research Assistant: April 2023 - Now
    • Computer Vision Lab at IUST under supervision of Dr. Mohammad Reza Mohammadi
    • Worked on anomaly localization techniques using Variational Autoencoders (VAEs) and analyzed attention maps as well as GradCAM (Gradient-weighted Class Activation Mapping) for precise localization.
    • Conducted researched on out-of-distribution detection methods, including post hoc techniques such as Null Space Analysis.
    • Contributed to advancing the field of machine learning by leveraging out-of-distribution detection methods to identify incorrect labels during the training phase, enhancing model robustness and accuracy.
    • Gained hands-on experience in deep learning frameworks and tools such as TensorFlow or PyTorch, enhancing proficiency in model development and experimentation.

Professional Experience

  • Software Developer Intern: June 2021 - September 2021
    • Company: Arman Rayan Sharif, Tehran
    • Collaborated with a team of developers in the development and enhancement of the company’s AI tool APIs, including image, speech, and text processing using Python and Django Rest Framework, offering these cutting-edge capabilities to customers.
    • Focused on the API’s key features, including: Invoice Handling, Throttling Mechanism, File Manager, Ticketing System and Request Logging.
    • Utilized Docker for containerization, enabling efficient deployment and management of applications and services.
  • Software Developer: September 2021 – December 2022
    • Company: Arman Rayan Sharif, Tehran
    • After finishing my internship, I got a return offer and transitioned from a successful software developer intern to a part-time software developer role at Arman Rayan Sharif, continuing to contribute to the company’s APIs and other software projects.
    • Expanded responsibilities to work on projects involving Apache Kafka for Real-time Data Streaming and Microservices Communication, Elasticsearch to implement advanced search and data retrieval capabilities, Grafana to create dashboards and visualization tools, providing real-time monitoring and insights into system performance and data metrics and deployed and managed applications using Kubernetes, leveraging container orchestration for scalable and resilient microservices architecture.

Teaching Experience

  • Computational Intelligence
    • Fall 2023
    • Instructor: Dr. Nasser Mozayani
  • Software Engineering
    • Fall 2023
    • Instructors: Dr. Behrouz Minaei-Bidgoli, Dr. Mehrdad Ashtiani
  • Internet Engineering
    • Fall 2022
    • Instructor: Dr. Vesal Hakami

Selected Projects

  • COVID-19 Prediction From Chest XRay ImagesDeep learning, Transfer learning, PytorchFall 2022
  • Tile Crack SegmentationDeep learning, Computer vision, Semantic segmentation, KerasFall 2022
    • Final project of IUST’s computer vision course.
    • Implementation of the U-Net model to segment cracks in tile pictures.
    • The challenge of this project was the design of the tiles, which had shapes that were similar to cracks.
  • Improving 5G Intrusion Detection with Synthetic Data and Transformer-Based ClassificationDeep Learning, Generative Adversarial Networks, Transformer, Unbalanced DatasetSpring 2023
    • This is my final thesis under the supervision of Dr. Vesal Hakami.
    • I address the issue of class imbalance in the existing datasets. I mainly focus on the 5G-NIDD dataset.
    • A conditional GAN model was used to generate data for the purpose of balancing the dataset.
    • I used the FT-Transformer architecture for classification and managed to improve the performance on the dataset compared to other methods.
  • IMDB Movie Genre ClassificationDeep Learning, NLP, KerasWinter 2022
    • Final project of IUST’s NLP course.
    • Multi-label classification of movie genres based on their plot.
    • The IMDB website was crawled in order to collect the dataset. Afterwards, preprocessing methods were used.
    • Three models were implemented for classification: Simple dense model, LSTM based model, CNN based model.

Technical Skills

  • Programming Languages: Python, Javascript, Java, C, Bash, Matlab
  • ML/DL Tools: Tensorflow, Keras, Pytorch, scikit-learn, OpenCV, Numpy, Pandas, Matplotlib
  • Web Development: Django, Flask, FastAPI, React, HTML, CSS
  • Devops/Databases: Docker, Kubernetes, PostgreSQL, Redis, Elasticsearch, Nginx
  • Others: Git, Linux, Latex, Grafana, Prometheus, Kibana, ANTLR, Selenium

Languages

  • Persian: Native
  • English: Fluent(IELTS: Overall: 8, Listening: 9, Reading: 9, Writing: 7, Speaking: 7)

Selected Courses

  • Deep Learning: A+
  • Computer Vision: A
  • Natural Language Processing: A
  • System Analysis and Design: A+
  • Software Engineering: A+
  • Data Structure: A+
  • Signals and Systems: A
  • Computer Security: A+
  • Database Design: A
  • Differential Equations: A+
  • Object Oriented Analysis and Design: A
  • Internet Engineering: A

Cetifiactes

References