Cognition.AI is an Artificial Intelligence Lab in the Department of Computer Science and Engineering at Manarat International University in Dhaka, Bangladesh. The lab is directed by Naimul Haque and has two research assistants, Md. Bodrul Alam and Abdullah Ath Towfiq. The lab's research focuses on several areas, including NLP, deep learning, machine learning, and computer vision. The lab has published several papers, including one on Bangla toxic comment classification and one on house rent prediction in Dhaka City. The lab's research assistants have completed several projects, such as automatic license plate recognition systems, colorization of gray-scale portraits, super-resolution of Bangla license plate images, and analysis of blockchain protocols. The lab also provides teaching support for undergraduate classes.

Expand...

Older news...

Here are just a few publications:

Gray-scale Portrait Colorization using Deep Learning

Automatic Colorization of a gray-scale image sounds trivial at first but it's not that simple. With the advance of AI and Deep Learning, full automation of colorization has been achieved only in recent years. In my undergrad thesis, I used Deep Learning techniques to color only portrait images using Convolutional Neural Networks (CNN) and a pretrained VGGFace descriptor. {View Project}

Bangla License Plate Restoration using a Super-Resolution Generative Adversarial

A promising approach of super-resolution, SR Generative Ad-versarial Networks, to upscale the low-resolution Bangla license plate images to super-resoltition in video surveillance systems.{View Project}

Bangla Handwritten Digits Datasets For Machine Learning

Compressed top 3 Bangla Handwritten digit datasets (NumtaDB BanglaLekha Ekush) for Machine Learning. {View Project}

Sales prediction with Random Forest

This is a {Kaggle Competition} based on a dataset consisting of daily sales data, provided by one of the largest Russian software firms (1C Company). The challenge is to predict the total sales for every product and store in the next month. I prepared a notebook demonstrating my skills in Machine Learning as well as Data Wrangling. {View Project}

Four Stage Automatic License Plate Recognition System for Bangla Low Resolution License Plate images

Automatic License Plate Recognition (ALPR). System is a crucial task that is used in numerous critical situations. Our method is intended to be effective not just on license plates with high resolution, but even when the plate (LP) is of very low resolution (LR).
{View Project}

Severity Rating and Classfication of Bangla Toxic Coomment using Deep Learning approaches

Automatic Colorization of a gray-scale image sounds trivial at first but it's not that simple. With the advance of AI and Deep Learning, full automation of colorization has been achieved only in recent years. In my undergrad thesis, I used Deep Learning techniques to color only portrait images using Convolutional Neural Networks (CNN) and a pretrained VGGFace descriptor. {View Project}

Super Resolution using Enhanced SR-Generative Adversarial Networks on Bangla License Plate Images

A promising approach of super-resolution, SR Generative Ad-versarial Networks, to upscale the low-resolution Bangla license plate images to super-resoltition in video surveillance systems.{View Project}

Access Control based on facemask detection using Opencv

Compressed top 3 Bangla Handwritten digit datasets (NumtaDB BanglaLekha Ekush) for Machine Learning. {View Project}

Bangla Sentiment Analysis using Natural Language Processing and Machine Learning

This is a {Kaggle Competition} based on a dataset consisting of daily sales data, provided by one of the largest Russian software firms (1C Company). The challenge is to predict the total sales for every product and store in the next month. I prepared a notebook demonstrating my skills in Machine Learning as well as Data Wrangling. {View Project}

Prediction of Covid-19 from Chest X-ray Images using CNN and pre-trained Models

A promising approach of super-resolution, SR Generative Ad-versarial Networks, to upscale the low-resolution Bangla license plate images to super-resoltition in video surveillance systems.{View Project}

Super Resolution using Enhanced SR-Generative Adversarial Networks on Bangla License Plate Images

A promising approach of super-resolution, SR Generative Ad-versarial Networks, to upscale the low-resolution Bangla license plate images to super-resoltition in video surveillance systems.{View Project}

Analysis of the Modified Consensus Protocols of Blockchain

A promising approach of super-resolution, SR Generative Ad-versarial Networks, to upscale the low-resolution Bangla license plate images to super-resoltition in video surveillance systems.{View Project}

Portrait Image Colorization using Generative Adversarial Networks

A promising approach of super-resolution, SR Generative Ad-versarial Networks, to upscale the low-resolution Bangla license plate images to super-resoltition in video surveillance systems.{View Project}

Current:

• Lecturer, April 2019 - Present, Department of CSE, Manarat Int. University

Responsibilities:

• Conducting undergraduate classes
• CSE-437: Computer Vision and Robotics
• CSE-433: Neural Network and Fuzzy Systems
• CSE-436: Pattern Recognition
• CSE-210: Algorithm
• CSE-104: Discrete Mathematics
• Undergraduate thesis supervision
• Preparing Course Offering and Academic Routines
• Training Competitive Programming