Banyar Oo

Banyar Oo

Web Dev × NLP

Education

University of Computer Studies, Yangon Bachelor of Computer Science (Knowledge Engineering)

Experience

METATEAM MYANMAR CO.,LTD.

Web Development Internship Trainee

[May 2026 - July 2026]

SOI ASIA EBA-KRI Kanchanaburi Fieldwork 2026

Chulalongkorn University & Mahidol University (Kanchanaburi Campus)

Participant Researcher

[11 - 17 June 2026]

  • Attended a pre-fieldwork online course on evidence-based approach, including data collection, interviewing, and data visualization techniques.
  • Participated in an international fieldwork program on resilient infrastructure and sustainable resource management under the theme "Blueprints for Resilience".
  • Engaged in data collection, site visits, and analysis related to disaster mitigation and water management.
  • Collaborated with students and researchers from Asia-Pacific region universities.

UCSY NLP Lab

Research Intern

[Nov 2025 - Sep 2026]

  • Conducted research on machine translation for Burmese-to-ethnic minority languages.
  • Preprocessed and curated parallel corpora for training translation models.
  • Trained, evaluated, and optimized neural machine translation models using techniques such as bidirectional training and LoRA-based fine-tuning.
  • Experimented with data augmentation techniques and hyperparameter tuning to improve model performance.

UCSY NLP Lab

Backend Developer

[Oct 2024 - Sep 2026]

  • Volunteered in the development of the Student Information System (SIS) for the University of Computer Studies, Yangon, using Bootstrap and Flask frameworks.
  • Developed backend logic for a range of university functions, including student and faculty data management, exam mark processing, and academic record generation.

Publications

B. Oo, H. M. Mo, and H. M. S. Naing, "A Study on Multilingual Burmese Translation with a Fine Tuned NLLB Model," 23rd IEEE International Conference on Computer Applications (IEEE-ICCA), Yangon, Myanmar, Feb. 26, 2026. Click Here

Projects

MCQ Generation using T5 & Distractor Generation using Retrieval-Augmented Generation

Question Generation
NLP
LLM
RAG
HF Transformers

A multiple-choice question generation system using a T5 Transformer model combined with retrieval-augmented generation for distractor generation. Fine-tuned with SciQ corpus and a custom domain-specific corpus created from a Cambridge Biology textbook and constructed a vector-based retrieval pipeline to generate semantically relevant distractors.

View

Air-written Burmese Alphabet and Numeral Dataset

Computer Vision
Dataset
Python
OpenCV

This dataset was created for an OCR system that can recognize air-written Burmese symbols. It includes all Burmese alphabets and numerals recorded in 83×84 resolution, with 30 images per character.

View View

Burmese Alphabet and Numeral OCR

Computer Vision
OCR
Python
OpenCV

An OCR system that recognizes air-written Burmese alphabets and numerals using webcam-based finger tracking and machine learning. Built a real-time model predicting Burmese symbols from gestures.

View

Student Information System (SIS)

Backend
Python
Flask
Bootstrap

A server-side rendered web application built with Flask. It provides functionalities like assigning students and teachers to class sections, calculating marks, managing courses, and generating academic records.

View