Image_of_Anan Ghosh

Anan Ghosh

Currently, I am a Computer Engineering graduate student at North South University, Bangladesh. I’m an enthusiastic software engineer that enjoys doing analytical work and has a keen interest in Machine learning , Computer Vision and Natural Language Processing . I’m now working in the fields of Computer Vision, Natural Language Processing and Biomedical-based research . To compensate for complexities and analytical work, I’m interested in devising better problem-solving approaches and trying out new technologies and tools as needed.

I participated in a lot of extracurricular activities in addition to my academic work. Throughout my academic years, I had the privilege of working for the NSU ACM as a Sub Executive in the Team Provision and the NSU Cultural Team Coordinator.

Working Experience

  • LEGO IO
  • Junior Developer Intern

    Rockville MD, USA

    March 2023 - Present

    Education

    Bachelor of Science in Computer Science and Engineering

    North South University, Bangladesh

    January 2018 - April 2022

    Research

    Emotional Reactions and Family Resilience During COVID-19 Lockdown Period among Bangladeshi Families

    This is a research project that focuses on the emotional emotions and family resilience of Bangladeshi families during the COVID-19 lockdown period. The purpose of the study is to learn more about how the structure of the family affects the emotional states of families during lockdown.

    Image synthesis with Normalizing Flows [Pending]

    We applied Normalizing flows to help produce additional synthetic X-Ray image samples of chest X-Rays that tested positive for COVID-19 for this study. We present our technique to creating synthetic pictures using normalizing flows, as well as the theory underlying normalizing flows and the overall outcomes, in this study.

    Heartisan: ECG Data Monitoring and Arrhythmia Prediction through Continuous Learning from Patient Data [On Going]

    An individual can keep track of their cardiac arrhythmia situations and ECG graphs. Data and predicted beats can be forwarded for validation. Each predicted beat may be verified by doctors. Depending on a set of criteria, the model is trained after a specified amount of time.

    Training Pruned Language Model [On Going]

    A more greener, resource-saving, and substantially smaller model may be accomplished as well as a bigger model by lowering storage needs and enhancing inference computing efficiency without losing comparative performance in distinct down-streaming tasks. This model can also be used as a few-shot learner, with the network’s sparsity rate trimmed up to 90%.Using the Lottery Ticket Hypothesis, this few shot learning model may be reduced without compromising performance on Bangla NLP tasks.

    Bangla Music Genre Classification Using NLP

    Due to the Bangla music industry's explosive growth, a vast number of Bangla songs are produced almost every day. Songs from various genres are produced on by a significant number of producers, lyricists, vocalists, and performers. Classical, folk, baul, Rabindra Sangeet, Nazrul Geeti, cinema music, rock music, and fusion music are among the genres of Bangla music that have seen the most success. Differentiating across musical genres is crucial for musical classification and knowledge representation. Although Bangla music looks to be extremely diverse on its own, there hasn't been any significant study utilizing NLP to classify the different musical genres that may be found in Bangla music.

    Projects

    Janao

    In a densely populated country like Bangladesh, local law enforcement institutions struggle to provide protection to the majority of the population. A user may easily report crimes to a secure database using the Flutter-based app, and the local law enforcement agency can then take necessary action based on the information.

    Document Management System

    The goal of this project is to digitize existing Department of Finance records and data. The converted digital format will be stored at a storage location determined by the Department of Finance, where it will be attached using a simple document system capable of search and retrieval

    Airline Management

    We will create a database for a random commercial airline, complete with staff, positions, and personal data/information, as part of this project. Varied plane types, itineraries, and different routes are appropriate for different classes. We can easily run an airline company if we have authority over this sort of data.

    Safe Education

    In this project, we are attempting to create programs that will secure the question paper. The question paper will be encrypted once it is uploaded into the program. Only those with authorization can use a password to gain access to this program. The question paper will then be decrypted and returned to its original form.

    Skills

    • Programming Languages
    • Python, C, C++, Java

    • Front-end
    • HTML, CSS3, Bootstrap

    • Database
    • MySQL, Firebase

    • Back-end
    • Django

    • Frameworks & Libraries
    • Keras, Tensorflow, PyTorch, Pandas, Matplotlib, OpenCV, Scikit learn

    • Tools
    • Git, GitHub, Slack, Trello, LaTeX

    Volunteer experience

  • NSU ACM Student Chapter
  • SUB EXECUTIVE

    Dhaka, Bangladesh

    November 2020 - February 2022

    IN-CHARGE

    Dhaka, Bangladesh

    January 2020 - January 2021

    Interests

    As well as conducting research and studying, I love to the guitar and exploring new environments in order to learn more about their cultures and heritages. Always exert a determined and passionate endeavor to keep up with technological advancement, experienced in a diverse range of social activities and consistently attempts to blend in with the team. Make a conscious effort to collaborate effectively with others as well as to constantly attempt to be helpful.

    Research Interests Area

  • Natural Language Processing, Computer Vision
  • Artificial Intelligence, Machine Learning, Deep Learing
  • Biomedical Engineering
  • Contact

  • Phone: +880 1521227058
  • Email: anan.ghosh@northsouth.edu
  • Personal Email: ananghosh15@gmail.com