Profile :

Honor Program in Computer Science, UIT-VNU.

Interests :

Machine Learning, Deep Learning, Computer Vision, Data Science


  • ID Card Digitization and Information Extraction
    • Examined to use CRAFT, PIXELLINK, CTPN for text detection module and VIETOCR for text recognition module
    • Created synthetic data for ID card images
    • Built a post processing module based on database retrieval to correct text error
    • Designed a web service to deploy the service
  • Script Identification in the wild(SIW2021) - ICDAR2021(A - Rank Conference)
    • Developed EfficientNet B7 for a competive solution
    • Ranked top 4 in the final leader board and become co-author in the associated conference paper in ICDAR 2021
  • Abdominal Multi Organ Segmentation (FLARE2021) - MICCAI2021(A - Rank Conference)
    • Presented an approach based on 2D segmentation
    • Exploit model optimization techniques in Pytorch to optimize infernece speed and GPU memory
    • Ranked top 6 in the final leader board and made an oral presentation in the satteline event of MICCAI conference
  • Vehicle Counting - Ho Chi Minh City AI-Challenge 2020
    • Designed a solution using Detectron2, DeepSort
    • Ranked top 18/217 in the final leader boardTraffic Sign Detection - Zalo AI Challenge 2020
    • Designed a solution using Detectron2
    • Ranked top 17 in the final leader board and get Honorable Mention Award
  • Smart Citizens for Smart Cities - AI Hackathon 2020 @ AI4VN
    • Creative Ideas Award
    • Honorable Mention Award
    • Awarded for the fastest and most efficient submit(3/10 submits)
  • Generative Adversarial Networks for Photo Cartoonization
    • Build an Style Tranfer network for photo cartoonization
    • Deploy the model into a mobile app
  • Landmark Retrieval
    • Experiment the performance of many Deep Leaning based feature extractors
    • Study the use of DELG model for the retrieval problem
    • Deploy a web application that can search for related image as well as their location information based on an inputimage
  • View Synthesis
    • Study the use of NERF model
    • Finish the completed pipeline for data formatting and model retraining


  • Das, A., Ferrer, M. A., Morales, A., Diaz, M., Pal, U., Impedovo, D., … & Gattal, A. (2021, September). ICDAR 2021 Competition on Script Identification in the Wild. In International Conference on Document Analysis and Recognition (pp. 738-753). Springer, Cham.

Contact :

E-mail :