AI intern
Resume:
QUANG-HUY PHAM
Bachelor in Data Science
HCM, VietNam
R ******@*****.***
github.com/qhnhynmm
SUMMARY
I am a person with the capacity and capability to han- dle immense pressure. I have a passion for data and AI. I have a passion for scientific research and learning new technologies and I have good teamwork skills
SKILLS
Programming:Python(proficient), C/C++, R, SQL(prior ex- perience).
Frameworks: PyTorch(proficient), TensorFlow, Keras, Spark(prior experience).
Tools: VSCode, Jupyter Notebook, Terminal,
Google Colab, Git, Docker, PowerBI.
EXPERIENCE
Research Student - University of Information and Technology - Vietnam National University HCM 06/2023 - Current Research Student - The UIT Natural Language Processing Group 01/2024 - Current Research Member - The UIT Data Science Society 05/2023 - Current
• Field: Multimodal (VQA), Natural Language Processing (NLP).
• Advisor: MSc. Kiet Van Nguyen, PhD. Trong-Hop Do. RESEARCH PROJECTS
Vietnamese AI Generated Detector
Paper research - Role: Leader - Team size: 4
• Tool & Technology: PyTorch, Natural Language Processing, Deep Learning, Data Crawling.
• Build new datasets and conduct deep learning models capable of distinguishing between human-written text and AI-generated text in the education domain.
• Publications: Accepted paper by ICIT2024 conference.
• Source code: [HERE]
ViTextVQA: A Large-Scale Visual Question Answering Dataset for Evaluating Vietnamese Text Comprehension in Images Paper research - Role: Leader - Team size: 4
• Tool & Technology: PyTorch, Multimodal, Visual Question Answering, Computer Vision, Natural Language Processing.
• Create the first high-quality large-scale dataset for text-based VQA task in Vietnamese, focusing on scene text and text appearing in the image. Through our extensive experiments, we found that VQA models using ViT5 as their back- bone behave as the answer selector methods when OCR text is suffixed for the question. Our experiments show the effectiveness of arranging from top-left to bottom-right, resulting in remarkable enhancements in the performance.
• Publications: In process review at Q1 journal.
• Source code: [HERE]
ViOCRVQA: Novel Benchmark Dataset and VisionReader for Visual Question Answering by Understanding Vietnamese Text in Images
Paper research - Role: Leader - Team size: 4
• Tool & Technology: PyTorch, Multimodal, Visual Question Answering, Computer Vision, Natural Language Processing.
• Create the largest-scale dataset for text-basedVQAtasksinVietnamese,focusing on text appearing in images. Through extensive testing, we found that the VQA models used for English are not really effective on Vietnamese. We recom- mend our proposed VisionReader model.
• Publications: In process review at Q1 journal.
• Source code: [HERE]
Pre-trained Language Models Fine-tuned with SVM for Legal Textual Entailment Recognition Paper research - Role: Leader - Team size: 4
• Tool & Technology: PyTorch, Language Model, Textual Entailment Recognition, Fine-tuned.
• The breakthroughs in natural language processing are not only a crucial step in technological evolution, but also yield significant benefits across various fields demanding high intelligence and precision. One of the notable applications is in the analysis and processing of legal text.
• Publications: Accepted paper by VLSP2023 workshop.
• Source code: [HERE]
COMPETITION PROJECT
VLSP 2023 Challenge on Visual Reading Comprehension for Vietnamese (10/2023 - 12/2023) Competition project - Role: Leader - Team size: 4
• Source code: [HERE]
UIT Data Science Challenge Group B - Vietnamese Fact Checking (09/2023 - 11/2023) Competition project - Role: Leader - Team size: 4
• Source code: [HERE]
Data Science Advanced Analysis 2023 Competition - Link Prediction for Wikipedia Articles (04/2023 - 06/2023) Competition project - Role: Leader - Team size: 4
• Source code: [HERE]
ACHIEVEMENTS
• Third prize - VLSP2023: Legal Textual Entailment Recognition Organized by VLSP - Vietnamese Language and Speech Processing 12/2023
• Top 4 private test - VLSP2023: Visual Reading Comprehension for Vietnamese Organized by VLSP - Vietnamese Language and Speech Processing 12/2023
• First prize - Group B DSC 2023: Vietnamese Fact Checking Organized by Faculty of Information Science and Engineering, University of Information Technology 11/2023
• Top 4 private test - DSAA 2023: Link Prediction for Wikipedia Articles Organized by Data Science Advanced Analysis Conference (rank A core 2021) 06/2023 PUBLICATIONS
• MAT: Effective Link Prediction via Mutual Attention Transformer Accepted by Data Science Advanced Analysis Conference (DSAA) (rank A core 2021)
• ViTextVQA: A Large-Scale Visual Question Answering Dataset for Evaluating Vietnamese Text Comprehension in Images
• ViOCRVQA: Novel Benchmark Dataset and VisionReader for Visual Question Answering by Understanding Viet- namese Text in Images
• Vietnamese AI Generated Detector
EDUCATION
University of Information and Technology - Vietnam National University HCM
• Bachelor in Data Science
• GPA: 8.3/10.
LANGUAGES
English - TOEIC 550
REFERENCES
• MSc. Kiet Van Nguyen
Vice Dean, Faculty of Information Science and Engineering University of Information and Technology
Email: ******@***.***.**
• PhD. Trong-Hop Do
Faculty of Information Science and Engineering
University of Information and Technology
Email: *****@***.***.**
Contact this candidate