Intro (Find your typing speed here!)
Experience
Apurba Technologies Ltd.
Software Engineer, Machine Learning
Sep 2022 - Present, Dhaka, Bangladesh
- My primary responsibilities in this job are designing, developing, researching and documenting Machine Learning based solutions.
Apurba‑NSU R&D Lab
Part-Time Research Assistant
Sep 2021 - Aug 2022, Dhaka, Bangladesh
- Published a paper which improves performance of OCR models on morphologically languages like Bangla.
- Worked with several NLP based projects.
ECE Department, North South University
Part-Time Lab Instructor
Spring 2022 & Summer 2022, Dhaka, Bangladesh
- Taught Database Management Systems (DBMS) using SQL and Object‑Oriented Programming (OOP) using Java to an average class size of 40 students.
- Prepared lab materials and provided exam questions; examined students’ answer scripts.
- Provided assessment and guidance to students as they worked on their projects.
Education
North South University
B.S. In Computer Science and Engineering
Jan 2017 - May 2021, Dhaka, Bangladesh
- Graduated with Magna Cum Laude distinction, CGPA ‑ 3.73/4.00
- Thesis title: Domain Adaptation on Speaker Recognition Problem with RawNet.
- Relevant Courses: Pattern Recognition and Neural Network, Natural Language Processing, Introduction to Linear Algebra.
BAF Shaheen College Kurmitola
HSC - Higher Secondary Certificate
2014 - 2016, Dhaka, Bangladesh
- GPA: 5.00/5.00
Rajoir Gopalgonj KJS Pilot Institution
SSC - Secondary School Certificate
2012 - 2014, Dhaka, Bangladesh
- GPA: 5.00/5.00.
Publications
LILA-BOTI : Leveraging Isolated Letter Accumulations By Ordering Teacher Insights for Bangla Handwriting Recognition. [ PDF ]
ICPR: International Conference on Pattern Recognition 2022
August 21-25, 2022 • Montréal Québec
- Abstract: Word-level handwritten optical character recognition (OCR) remains a challenge for morphologically rich languages like Bangla. The complexity arises from the existence of a large number of alphabets, the presence of several diacritic forms, and the appearance of complex conjuncts. The difficulty is exacerbated by the fact that some graphemes occur infrequently but remain indispensable, so addressing the class imbalance is required for satisfactory results. This paper addresses this issue by introducing two knowledge distillation methods: Leveraging Isolated Letter Accumulations By Ordering Teacher Insights (LILA-BOTI) and Super Teacher LILA-BOTI. In both cases, a Convolutional Recurrent Neural Network (CRNN) student model is trained with the dark knowledge gained from a printed isolated character recognition teacher model. We conducted inter-dataset testing on BN-HTRd and BanglaWriting as our evaluation protocol, thus setting up a challenging problem where the results would better reflect the performance on unseen data. Our evaluations achieved up to a 3.5% increase in the F1-Macro score for the minor classes and up to 4.5% increase in our overall word recognition rate when compared with the base model (No KD) and conventional KD.
Skills
| Programming | Proficient in Python, C, C++ and Java. Knows C#, PHP and Javascript. |
| Development | Proficient in Clean Coding concepts and Object Oriented Design Patterns |
| Machine Learning | Experience working with PyTorch, Keras, scikit‑learn, pandas, pyspark and matplotlib. |
| Database | Intermediate to advanced level Database Management System knowledge using MySQL. |
| Version Controlling | Well‑read on the best practices of Version Controlling using Git |
| Communications | A skilled communicator who is comfortable speaking in public and writing papers. |
| Languages | Bangla, English and Italian. |
Certifications
| Certificate | Provider | Issue Date |
|---|---|---|
| Deep Learning Specialization | Coursera | November 29, 2020 |
| Machine Learning | Coursera | July 21, 2020 |
| Introduction to Philosophy | Coursera | August 8, 2020 |
Achievements
Robi Datathon 2.0
Finalists
Robi, Axiata Ltd., 2022, Dhaka, Bangladesh
- Among top 25 out of 384 teams.
- Plotted geographic maps using geospatial data.
- Optimized pyspark queries for faster queries on large data.
- Implemented cost sensitive learning to bump performance of a machine learning model.
MIST Inter‑University ICT Innovation Fest Hackathon
Champions
MIST, 2021, Dhaka, Bangladesh
- Deployed a handwriting recognition model.
- Designed a handwriting identification model.
IOT For Tomorrow
First Runners Up
NSU IEEE Student Branch, 2019, Dhaka, Bangladesh
- Presented a fire detection circuit using a microcontroller integrated with Firebase.
Electrathon
2nd Runners Up
NSU IEEE Student Branch, 2018, Dhaka, Bangladesh
- Written a python automation script using selenium.
- Built an Arduino solution to the given problem
National Talent Hunt
Best Talent in Computer and Math in Madaripur
Bangladesh Education Ministry, 2012, Dhaka, Bangladesh
- Participated in a series of competitive tests on Mathematics and Computer Science up to District level.
Projects
North South University
B.S. In Computer Science and Engineering
Jan 2017 - May 2021, Dhaka, Bangladesh
- Graduated with Magna Cum Laude distinction, CGPA ‑ 3.73/4.00
- Thesis title: Domain Adaptation on Speaker Recognition Problem with RawNet.
- Relevant Courses: Pattern Recognition and Neural Network, Natural Language Processing, Introduction to Linear Algebra.
Extracurricular
North South University
B.S. In Computer Science and Engineering
Jan 2017 - May 2021, Dhaka, Bangladesh
- Graduated with Magna Cum Laude distinction, CGPA ‑ 3.73/4.00
- Thesis title: Domain Adaptation on Speaker Recognition Problem with RawNet.
- Relevant Courses: Pattern Recognition and Neural Network, Natural Language Processing, Introduction to Linear Algebra.
Interests
North South University
B.S. In Computer Science and Engineering
Jan 2017 - May 2021, Dhaka, Bangladesh
- Graduated with Magna Cum Laude distinction, CGPA ‑ 3.73/4.00
- Thesis title: Domain Adaptation on Speaker Recognition Problem with RawNet.
- Relevant Courses: Pattern Recognition and Neural Network, Natural Language Processing, Introduction to Linear Algebra.