Su Myat Noe, Ph.D.

Researcher in Multimodal Generative AI & Vision-Language Models

About Me

A multidisciplinary researcher in Computer Science with expertise in Computer Vision, Artificial Intelligence, Machine Learning, and Deep Learning. I recently completed a Ph.D. at the University of Miyazaki, Japan, with a research focus on developing deep learning-based methods for detecting estrus behavior in cattle using multi-camera tracking systems.

Currently serving as a Project Researcher at the National Institute of Informatics (NII) in Tokyo, Japan, I contribute to advanced research in vision-language modeling, retrieval-augmented generation (RAG) systems, agentic AI, video question answering (VideoQA), and video summarization. My research contributions have been published in leading international conferences and journals, with the aim of advancing impactful and practical AI solutions for real-world applications.

Beyond academia, I actively engage in leadership and advocacy roles within the AI community. As the Ambassador of Women in AI Myanmar, I work tirelessly to promote gender inclusivity and diversity in the AI domain. Additionally, serving as the Google Developer Student Club Lead of Japan 2022-2023 allows me to connect with fellow AI enthusiasts, providing them with valuable learning opportunities and resources. I find great joy in organizing and speaking at tech events and conferences worldwide, where I share my knowledge and experiences, inspiring others to pursue their aspirations in AI.

Experience & Activities

Professional Experience

Project Researcher

1 Apr 2025 – Present

National Institute of Informatics (NII), Tokyo, Japan

  • Conduct advanced research on multimodal large language models (LLMs) that integrate image, video, and natural language understanding, with applications in generative AI for the Asia-Pacific (APAC) region.
  • Design, implement, and evaluate culturally-adapted image and video generation pipelines using diffusion models, autoregressive generative models, and LoRA-based fine-tuning techniques.
  • Develop Retrieval-Augmented Generation (RAG) systems to enhance multimodal content retrieval and Video Question Answering (VideoQA) capabilities.
  • Build and curate large-scale, high-quality datasets for training and fine-tuning vision-language generative models, ensuring diversity and cultural relevance in generated outputs.
  • Collaborate with interdisciplinary teams across research and product groups to drive innovations from concept to prototype, focusing on scalable and impactful generative AI solutions.
  • Work closely with Ethics and Governance teams to ensure generative outputs meet ethical standards, uphold safety guidelines, and provide broad societal benefits.

Project Activities

Detection of Lameness Behavior in Cattle (Lab Project)

2022 – 2023

Project led by the Ministry of Internal Affairs and Communications, Japan

  • Core member of a significant project focusing on utilizing local 5G networks for practical problem-solving.
  • This initiative aimed to develop a real-time detection and tracking system for lameness behavior in cattle.
  • My core responsibilities included implementing a computer vision program to detect and track cattle lameness behavior using advanced deep-learning techniques, then testing our system on the real-time 5G network to check its performance.

Volunteering & Community Leadership

Women in AI Myanmar, Ambassador

1 Dec 2020 – Present

My focus is empowering women to take leadership positions within AI and machine learning industries while also creating opportunities for underprivileged youth, particularly girls, to gain valuable skills and knowledge in this rapidly evolving field.

Google Developer Student Club Lead, Japan

1 Aug 2022 – Aug 2023

Through the Google Developer Student Club (GDSC) network, I spearheaded initiatives to empower next-generation undergraduate students to innovate through emerging AI technologies by introducing engaging campus activities that fostered hands-on learning.

Internship Experience

Bagan Innovation Technology, Junior Data Scientist

1 Sep 2018 – 30 Nov 2018
  • Collaboratively built Natural Language Processing (NLP) models for the Myanmar Language that work with limited resource settings.
  • Performed text mining and web scraping from online data sources using Python libraries.
  • Developed data pipelines and machine learning models for Sentiment Analysis for Myanmar Language to be used in chatbot offerings.

Fujitsu Japan, Junior System Developer

1 Oct 2017 – 30 Nov 2017
  • Performed day-to-day duties such as testing system functionalities to ensure the software quality.
  • Developed a Hotel Reservation System with PHP, JavaScript, and jQuery Language by following the fast-paced agile development cycle.
  • Produced source codes by converting logical sequences and workflows from the project requirements.

Invited Talks

  • 2025 Aug: Women in AI Myanmar (Virtually), Machine Learning Summer School, “Vision Meets Language: Ensuring Safety in Multimodal Video AI”
  • 2025 Aug: Google I/O Extended Yangon 2025 (Virtually), “Empowering Safer Agents: The Future of Vision-Language AI”
  • 2023 Mar: The Second Sakura Workshop, Taichung Taiwan, “A Study on Automatic Black Cattle Tracking with Computer Vision and Deep Learning”
  • 2022 Nov: Google Developer Expert, DevFest, Yangon, Myanmar (Virtually), “Object Detection and Instance Segmentation with Detectron 2 Framework”
  • 2022 Jul: Google Developer Expert, DevFest, Yangon, Myanmar (Virtually), “Knowledge Sharing About Computer Vision Kaggle”
  • 2019 Jul: Technical Workshop at University of Information Technology Myanmar, “How to Build a Machine Learning Community in a Developing Country”

Education

Ph.D. in Engineering

1 Apr 2022 – 25 Mar 2025

University of Miyazaki, Japan

Department of Materials and Informatics, Computer Science and Bioinformatics Course, Visual Information Processing Lab. Advisor: Dr. Thi Thi Zin.
Research focused on deep learning-based detection and tracking of animal behavior using multi-camera systems, with transferable methods applicable to multimodal AI and large-scale generative models.

M.Eng. in Engineering

1 Apr 2020 – 31 Mar 2022

University of Miyazaki, Japan

Department of Energy and Electronics, Visual Information Processing Lab. Advisor: Dr. Thi Thi Zin.

Research Student

1 Oct 2019 – 31 Mar 2020

University of Miyazaki, Japan

B.C.Sc. in Computer Science

1 Dec 2013 – 31 Dec 2018

University of Information Technology (UIT), Myanmar

Major of Knowledge Engineering.

High School

1 July 2003 – June 2013

Practicing High School Yangon Institute of Education, Myanmar

Publications & Thesis

Peer-Reviewed Journals

  • Optimizing black cattle tracking in complex open ranch environments using YOLOv8 embedded multi-camera system.
    Su Myat Noe, Thi Thi Zin, Pyke Tin, and Ikuo Kobayashi.
    Scientific Reports, 15, Article no. 6820. (2025)
  • Precision Livestock Tracking: Advancements in Black Cattle Monitoring for Sustainable Agriculture.
    Su Myat Noe, Thi Thi Zin, Pyke Tin, and I. Kobayashi.
    Journal of Signal Processing , 28(4), pp. 179-182. (2024)
  • Comparing State-of-the-Art Deep Learning Algorithms for the Automated Detection and Tracking of Black Cattle.
    Su Myat Noe, Thi Thi Zin, Pyke Tin and Kobayashi, I. Kobayashi.
    Sensors, 23(1), p.532. (2023)
  • Automatic detection and tracking of mounting behavior in cattle using a deep learning-based instance segmentation model.
    Su Myat Noe, Thi Thi Zin, Pyke Tin, and I. Kobayashi.
    International Journal of Innovative Computing, Information, and Control , 18(1). (2022)
  • BIS Reasoning 1.0:The First Large-Scale Japanese Benchmark for Belief-Inconsistent Syllogistic Reasoning. In: Proceedings of the 2025 Conference on Computational Linguistics.
    Nguyen, H.-T., Liu, C., Takeda, K., Miyao, Y., Stenetorp, P., Liu, Q., Su Myat Noe and Tachibana, H., 2025.
    In: Proceedings of the 2025 Conference on Computational Linguistics. [online] Available at: https://example.org/bis-reasoning-1.0 [Accessed 22 July 2025].

Peer-Reviewed Conference Proceedings

  • Su Myat Noe, et al. “From Vision to Vocabulary: A Multimodal Approach to Detect and Track Black Cattle Behaviors”, The 16th International Conference on Genetic and Evolutionary Computing (ICGEC).
  • Su Myat Noe, et al. “Precision Livestock Tracking...”, 2024 RISP International Workshop on NCSP.
  • Su Myat Noe, et al. "Efficient Segment-Anything Model for Automatic Mask Region Extraction...", IEEE ICCE-Berlin 2023.
  • Su Myat Noe, et al. "A Deep Learning-based Solution to Cattle Region Extraction for Lameness Detection", 2022 IEEE LifeTech.
  • Thi Thi Zin, Su Myat Noe, et al. "An Intelligent Method for Detecting Lameness in Modern Dairy Industry", 2022 IEEE LifeTech.
  • Su Myat Noe, et al. “A Deep Learning-based Instance Segmentation Model using Mask-RCNN...”, ICICIC2021.
  • Su Myat Noe, et al. “Automatic Detection of Mounting Behavior in Cattle...”, IEEE LifeTech 2021. (IEEE WIE Student Paper Award)
  • Su Myat Noe, et al. “Detection of Estrus in Cattle by using Image Technology and Machine Learning Methods”, IEEE GCCE 2020.

Thesis

  • Doctoral Thesis (2025): "A Study on Deep Learning-based Automated Detection and Tracking of Estrus Behavior in Black Cattle"
  • Master’s Thesis (2022):“Cow Estrus Detection and Tracking based on Image Technology with the Enforcement of Deep Learning Methods”
  • Bachelor’s Thesis (2018): “Question Answering System using Bi-Directional Attention Flow”

Awards & Honors

  • 2025 Encouragement Women Research Award, University of Miyazaki, Japan
  • 2024 Best Student Paper Award, ICGEC Conference, Miyazaki, Japan
  • 2024-2025 University of Miyazaki, President’s Research Funding Award
  • 2024 Best Student Paper Award, RISP International Workshop on NCSP, Honolulu, Hawaii (Top – 1 out of 25 candidates)
  • 2022 Ministry of Japan Science Technology (JST) Researchers Grant for Doctoral Studies (2022-2025)
  • 2022 NEC C&C Non-Japanese Foreign Researchers Grant for Doctoral (2022-2023)
  • 2022 Scholarship, JASSO Scholarship for International Student, Japan
  • 2021 Women in Engineering Best Paper Award, 2021 IEEE 3rd Global Conference on LifeTech, Japan
  • 2020 Full Scholarship, Naoji Iwatani Full Scholarship for Master Program, Japan
  • 2019 Scholarship, JASSO Scholarship for International Students, Japan
  • 2019 Scholarship, Southeast Asia Machine Learning Summer School (SEAMLS), Indonesia

Technical Skills

Programming

  • Python, C++, MATLAB

Deep Learning Skills

  • TensorFlow, Keras, CUDA
  • Machine Learning, Deep Learning
  • Data/image processing algorithms
  • Automatic Dataset Generation
  • Building Datasets from Scratch
  • Supervised & Unsupervised Learning

Generative AI

  • Diffusion Models (Stable Diffusion, etc.)
  • Autoregressive Generative Models
  • Multimodal LLM Training & Evaluation

Computer Vision Skills

  • OpenCV, Image Processing, CNN Models

Libraries

  • Pandas, Numpy, Matplotlib, Keras, OpenCV

Version Control & Cloud

  • Git, Docker
  • Google Cloud, AWS

Additional Skills

  • Data Pipelines, NLP, Sentiment Analysis
  • Visualization, IoT, Network Simulation
  • High-Performance Computing

Language Skills

  • English (TOEFL IBT 81, TOEIC 850)
  • Japanese (JLPT N3)
  • Myanmar (Native)