Lecturer, Department of Computer Science and Engineering (CSE)
Asian University of Bangladesh (AUB), Dhaka, Bangladesh
Research Interest: Deep Learning, Computer Vision, NLP
Research Projects
A Real-time Bangla Sign Language Translation System
Working on a project to develop a pose-based Bangla sign language translation system to help deaf
people for effective communication. A novel transformer architecture has been proposed to model
temporal descriptors of the sign videos to Bangla sentence. A short version of this work has been
won in Artificial Intelligence for Bangla Competition 2.0.
Real-time Detection and Recognition of Bangladeshi License Plate
Working on a project to develop a real-time automatic license plate recognition (ALPR) system for
Bangladeshi vehicles.
Bangla Social Network Post Analysis using Deep learning
Working on a project to develop a real-time language model which is able to categorize ruling government post and anti-government post along with toxic level. For this work, we have collected a
dataset and then using this dataset we plan apply a transformer to build the model.
Real-time Sexual Harassment Detection in Workplace
Working on a project to develop a real-time automatic sexual harassment detection in workplace. A
short version of this work has been published in an IEEE conference 2021.
Automatic Detection of Unfit Vehicles for Intelligent Transportation System
An optimized YOLOv3-based unfit vehicles detection system where Mask R-CNN has been implemented
to detect the damaged portion of the vehicles This work has been submitted for peer review.
This work has been submitted for peer review
Garbage Classification using object detection algorithms
Working on garbage classification using state-of-the-art object detection algorithms. With the increasing concerns surrounding environmental sustainability, proper waste management has become a global priority.
Our project focuses on leveraging advanced computer vision techniques to automatically identify and categorize different types of waste items.
Automatic Detection of Unfit Vehicles for Intelligent Transportation System
An optimized YOLOv3-based unfit vehicles detection system where Mask R-CNN has been implemented
to detect the damaged portion of the vehicles This work has been submitted for peer review.
This work has been submitted for peer review
Exam Cheat Recognition using CNN
working on the development of a Convolutional Neural Network (CNN) based model specifically tailored
for the vital task of recognizing potential instances of cheating in offline exams.
Leveraging the power of deep learning, our model aims to analyze scanned answer sheets
and identify patterns that might indicate unauthorized collaboration or information access
during the examination process. Code
Research Projects 