Research Interest

• Deep Learning

• Natural Language Processing

• Computational Linguistics

• Computer Vision

• Image Processing

• Human Computer Interaction 

Project 1

Emotion Recognition from Bengali-English Code-Mixed text using Deep learning. 

Language and Frameworks: Python, scikit-Learn, TensorFlow, NLTK, Keras, NumPy, Pandas, Matplotlib. 

Emotion detection is a computational approach for finding the distinct emotion or feeling of an individual. Although Bengali is a low-resource language, the amount of Bengali-English codemixed textual data has grown significantly because of the recent widespread use of social media applications among Bengali users. Gradually, the classification of emotions in Bengali-English codemixed data has become a crucial challenge for applications in e-commerce, healthcare, suicidal attempt reduction and crime detection. Nevertheless, the lack of Bengali language processing techniques and Bengali-English dataset have made the emotion recognition more challenging. This research work offers a Deep Learning based approach for classifying emotions from Bengali-English code-mixed data into six basic categories: disgust, sadness, joy, anger, fear, and surprise. Due to the lack of required dataset, a Bengali-English code-mixed corpus consisting of 10,221 sentences is created. In order to identify the best features, this work investigates several word embedding techniques, including Word2Vec, FastText, and Keras Embedding Layer. Different types of of Machine Learning and Deep Learning based algorithms including the proposed technique using Word2Vec and BiLSTM are applied on the developed corpus. In order to find out the best technique, a comparative analysis among all the methods is demonstrated revealing that the BiLSTM with Word2Vec word embedding technique outperforms rest other models achieving the highest accuracy of 76.1%. 

Project 2

Sentiment Analysis from Twitter Data Data using Deep Learning. 

Language and Frameworks: Python, scikit-learn, NumPy, Pandas, NLTK, Matplotlib. 

Online marketing and e-commerce companies are booming in Bangladesh in this age of internet technology. As more people were afflicted with the COVID-19 epidemic, internet purchasing became the primary channel for closure shopping and was considered the safest method. The enterprises were pushed to appear online. There are many online service providers, such beneficial for individuals, but it also calls into question the quality of the products with services. Therefore, it is simple for new clients to be deceived, when doing internet purchasing. The enormous volume of tech gadget review data that is generated online every day can be examined for the purpose of assessing public sentiment and assisting in market intelligence. While the study of sentiment classification has advanced greatly in languages with abundant resources, it is still in the preliminary stage for languages with limited resources, such as Bengali. This work proposes a model for classifying the sentiment on online Bengali tech gadget reviews into three basic categories- positive, negative, and neutral. For this purpose, around 6015 Bengali tech review data is collected. Various Machine Learning techniques are then applied along with different feature extraction techniques. After evaluating the performance, the Random Forest outperforms the rest of other techniques, having a maximum accuracy of 86.28%. 

Project 3

Emotion Detection from Tamil Texts using Transformers

Language and Frameworks: Python, scikit-learn, NumPy, Pandas, NLTK, Matplotlib. 

Recently, emotion analysis has gained increased attention by NLP researchers due to its various applications in opinion mining, e-commerce, comprehensive search, healthcare, personalized recommendations and online education. Developing an intelligent emotion analysis model is challenging in resource-constrained languages like Tamil. Therefore a shared task is organized to identify the underlying emotion of a given comment expressed in the Tamil language. The paper presents our approach to classifying the textual emotion in Tamil into 11 classes: ambiguous, anger, anticipation, disgust, fear, joy, love, neutral, sadness, surprise and trust. We investigated various machine learning (LR, DT, MNB, SVM), deep learning (CNN, LSTM, BiLSTM) and transformer-based models (Multilingual-BERT, XLM-R). Results reveal that the XLM-R model outdoes all other models by acquiring the highest macro f1-score (0.33)

Project 4

Humanoid Robot Drawing and Animation in Computer Graphics 

Language and Frameworks: GLUT (OpenGL Utility Toolkit)

The goal of this project was to develop a 3D model of a humanoid robot using simple geometric forms and to display walking and waving animations. The robot’s bodily sections are represented by cylinders and a head represented by a sphere. 

Learning:

• Implementing geometric shapes (cylinders and spheres) to create a humanoid structure.

• Adding transformation methods for movement and rotation.

• Providing input options for animation through mouse click and keyboard.

• Creating animations by manipulating angles and joint connections.

• Experimenting with lighting and shading effects to enhance the visual of the figure.

• Learning to use GLUT to create interactive graphics applications.

Project 5

Image Manipulation, Sharpening and Stitching Project 

Language and Frameworks: Python, OpenCV 

The goal of this project is to learn Python and OpenCV-based image processing techniques, such as image manipulation, picture sharpening, and image stitching. We implemented image manipulation operations on raw images, after that focused enhancing image sharpness and in final module stitched together two input photos to produce a panoramic image. One of our motivations was to gain practical experience of working with image data, from straightforward manipulation to sophisticated image processing methods.  

Learning:

•   Reading and processing raw image data and converting color images to grayscale

•  Understanding and implementing image thresholding.

• Experimenting with image sharpening techniques

• Gaining familiarity with argparse for command-line parsing

•  Employing feature detection and matching using OpenCV

• Understanding key point extraction and description using SIFT and SURF

• Learning to develop a reusable class for image stitching