Navid Rajabi

I'm a PhD Candidate in Computer Science (AI/ML) at George Mason University and I have been lucky to be advised by Prof. Jana Kosecka.

My research lies at the intersection of Natural Language Processing (NLP) using Large Language Models (LLMs), Computer Vision, and Robotics/Navigation, aiming for efficient adaptation and development of Multimodal Large Language Models (MLLMs).

More specifically, my research focuses on fine-grained concepts understanding (e.g. grounding and spatial reasoning). Ultimately, my work aims to improve the Embodied AI agents performance for instruction following in photo-realistic 3D environments.

Prior to joining George Mason University, I was a graduate student at the University of San Francisco and Illinois State University , where I received my M.Sc. degree. During my master's degree, I was mainly working on ML and Deep Learning for the Internet of Things (IoT) robustness.

Email  /  LinkedIn /  Google Scholar  /  Github /  Twitter

profile photo
๐Ÿ”ฅ News
Research Projects
Vision-and-Language Navigation Improvement in Previously Unseen 3D Environments

The objective of my main research project is to enhance the Success Rate Weighted by Path Length (SPL) of the embodied agent in unseen environments by boosting the (1) language grounding and (2) planning/navigation components.

Performance Analysis on Image Captioning Models (Transformer-based/Language-side Focus)

In this project, we measure the effect of employing BERT-based generated embeddings and substituting the LSTM with a Transformer on the Image Captioning task. As a result, we achieved 13.5% higher performance (in terms of BLEU-4) compared to the baseline and noticed Out-of-Distribution object classes/words among the generated captions, due to the usage of BERT.

Selected ML Projects
Deep Learning Interpretation of Patients Chest X-Ray for Classification/Diagnosis on Stanford Benchmark (2019)

Top-Performing Setting: Google Xception (Accuracy = 84%, F-1 score = 68%)
Lightweight Setting: Google MobileNetV2 (Accuracy = 82%, F-1 score = 67%)

In this project, we only considered frontal studies and measured the performance of Deep Convolutional Neural Networks (pre-trained on ImageNet) for multi-class (14) binary classification as diagnosis (e.g. Pneumonia, Lung Opacity, Fracture etc).

In addition, we tried to implement the entire pipeline using ML approaches (i.e. without utilizing ConvNets for feature extraction) to improve the interpretability of the model. To be more specific, an unsupervised Stacked Auto Encoder (SAE) was trained on the images and the middle layer was taken as a latent embedding. Then, this embedding was fed into a Random Forest classifier. However, the performance was far behind the Deep Learning approach.

Selected Publications
Multi-Layer Perceptron Artificial Neural Network Based IoT Botnet Traffic Classification
Yousra Javed, Navid Rajabi
28th USENIX Security Symposium, Santa Clara, California, USA, 2019
Student Grant Awardee
[USENIX][Springer Nature]

The core competence of this work stems from the fact that this model has been trained, validated and tested on more than 7 million instances of an actual IoT network traffic (consisting of 9 real IoT devices), instead of simulated/synthetic traffic datasets.

A Shortcut for Caret Positioning on Touch-Screen Phones
Jianwei Lai, Navid Rajabi, Elahe Javadi
MobileHCI '19: Proceedings of the 21st International Conference on Human-Computer Interaction with Mobile Devices and Services, Taipei Taiwan, October 2019
[ACM]

We've proposed a more efficient shortcut to interact with the large-screen smart phones with one hand, which outperforms two of current approaches, in both quantitative and qualitative metrics. [NSF Grant]

LSTM-RNN Model for Voltage Abnormality Detection at IoT Gateway
Jihad Qaddour, Navid Rajabi
International Journal of Computer Applications, October 2019
[Paper][Dataset]

The main objective of this project is to detect abnormalities at the edge as a security-oriented proactive approach to recognize the future abnormalities in time-series voltage dataset of a given household.

SDIoBoT: A Software-Defined Internet of Blockchains of Things Model
Navid Rajabi, Jihad Qaddour
International Journal of Internet of Things, 2019
[Paper]

The main idea behind this model is to behave the transactions of a local IoT network as a distributed ledger of Blockchain by employing the Elliptic Curve Digital Signature Algorithm (ECDSA), which is being used in the Bitcoin architecture for transactions integrity.

Filter Bubble, Selective Exposure, and Integrative Complexity
Elahe Javadi, Nancy L. Novotny, Elnaz Mirrahimi, Navid Rajabi
MWAIS '17, Springfield, Illinois, USA, 2017
[Best Paper Award Winner]

Selective Exposure is a destructive phenomenon occurs due to the over usage of Recommender Systems. In this work, we've demonstrated the existence of this negative effect using the real, proprietary dataset we've collected from students.

Full Stack Projects
An End-to-End, Customized Data Analysis & Visualization Pipeline for A Local Startup (2019)
Role: Team Lead, Algorithm Designer, and Full Stack Developer
Tech Stack: Python, Scikit-learn, Pandas, Numpy, Matplotlib
Description: Saving about one week of the data analyst's work load per month (reduced to around 10 minutes) by fully automating the monthly data processing, as well as generating business-oriented reports for the executives (CEO and CFO).
A General-purpose, Quiz-taking Web App Platform with Randomness Feature for Question Selection (2019)
Role: Team Lead & Full Stack Developer
Tech Stack: Python (Flask), Gunicorn WSGi, Heroku, Apache Ubuntu, MySQL/SQLite, Bootstrap, Java Script, HTML, CSS
An Expedia-like Web App (2016)
Role: Full Stack Developer
Tech Stack: Java Servlet, Jetty, Apache Ubuntu, Google Map API, MySQL, Bootstrap, HTML, CSS, Java Script

Website design courtesy: Jon Barron