Artificial Intelligence Machine Learning

Tran Ngoc Xuan Tin

AI ML Engineer Fresher

Date of birth: 20/03/2001

Location: Vinhome Grand Park / District 9 / Ho Chi Minh City

§ TinAiworkspace ï TranTin

# ************@*****.*** H +84.08.329.04954

Summary

My name is Tran Ngoc Xuan Tin, a recent graduate from FPT University with a major in Artificial Intelligence (AI). Throughout my studies, I have acquired extensive knowledge in various AI-related fields such as machine learning, deep learning, natural language processing, and computer vision. I have participated in numerous practical projects and research activities, through which I developed my programming, data analysis, and problem-solving skills. I aspire to work at a growing company in the artificial intelligence industry, where I can learn and develop my skills while contributing to the company’s growth. In the near future, I hope to achieve the following goals:

– Accumulate extensive work experience in the field of research, implementation, and development of machine learning models and systems with artificial intelligence applications.

– Participate in research and development projects of Artificial Intelligence application technologies. I believe that with a passion and commitment to applying artificial intelligence to real-world applications, I am always eager to learn and embrace the latest technologies. I am confident that the skills and knowledge I have gained will enable me to contribute effectively to the development of the company and the AI industry. Education

11/2019 - 4/2024 Bachelor of Science in Artificial Intelligence Major in Artificial Intelligence, FPT University, Ho Chi Minh City. Subjects: Algebra, Data Structures and Algorithms, Object-Oriented Programming, Machine Learning, Deep Learning, Computer Vision, Natural Language Processing. GPA: 7.18/10 - Degree classification: Good. Link to evidence Projects

Virtual Shirt Fitting Graduation project Link to Demo

– This project is an image-based virtual experimental field that attempts to create realistic images of individuals in specific clothing items. Traditional approaches solve this task independently, curve the item to fit the person’s body, and create a division map of individual wear.

+ Researched existing models and selected the best model for the task.

+ Studied the given sample data and models.

+ Recreated input data for the model by processing original data through models like Densepose, Openpose, and Human- Parse v3(CIHP-v3).

+ Extracted regions of interest (i.e., clothing areas) after successfully recreating input images for the model.

+ Researched and optimized the existing model by analyzing and studying the layers within the network.

+ Trained the model with the best weights and deployed the model on the web through Anvil.

+ Documented the process and results in the final report.

– Techstack: GANs, OpenCV, Deep Learning, Python, Computer Vision, Pytorch, Tensorflow, Openpose, Densepose, CIHP-PGN, Image Processing, Anvil.

– Detailed results and information are included in the report file attached in the GitHub link. Vietnamese Sentiment Analysis Subject project Link to Demo

– Sentiment analysis, the task of automatically determining the sentiment expressed in a given text, plays a crucial role in various natural language processing applications. In recent years, fine-tuning large language models (LLMs) has emerged as a powerful approach for achieving state-of-the-art performance in sentiment analysis tasks.

+ Collected data from social networking sites (Facebook, Twitter, etc.) about a specific topic. Cleaned the data, removed unnecessary characters, and handled special cases such as emotional icons and abbreviations.

+ Applied text preprocessing techniques to clean and normalize data.

+ Developed and trained machine learning models for sentiment classification, utilizing a variety of architectures including RNN, RCNN, LSTM, BERT, PhoBERT, and GPT-2.

+ Evaluated and compared the performance of each model on the preprocessed dataset, considering metrics such as accuracy, sentiment accuracy, label accuracy, and other relevant measures.

– Techstack: Nature Language Processing, Deep Learning, RNN, LSTM, GRU, Bert, PhoBert, GPT-2.

– Detailed results and information are included in the report file attached in the GitHub link. Vietnamese License Plate Recognition Subject project Link to Demo

– Vehicle license plate recognition is an important technology for traffic management, safety, and security. In Vietnam, as in many other countries, vehicles are required to display unique license plates that are used to identify them. Automated license plate recognition systems use computer vision techniques to extract the license plate number from images and videos of vehicles.

+ Gathered and processed data consisting of images of license plates and individual alphanumeric characters found on license plates.

+ Ensured the quality of the labelled data through rigorous verification processes.

+ Employed YOLOv7 model for license plate detection in images or videos, fine-tuning hyperparameters to achieve optimal performance.

+ Utilized EasyOCR model or YOLOv7( Trained with data that are individual alphanumeric characters found on license plates ) for optical character recognition (OCR) to read the license plate numbers extracted from the images.

+ Combined the recognized license plate numbers with the original images or videos to obtain the final results.

+ Compiled a detailed report summarizing the methodology, results, and findings of the project.

– Techstack: OpenCV, Deep Learning, Python, Computer Vision, Pytorch, Yolo, EasyOCR, Image Processing.

– Detailed results and information are included in the report file attached in the GitHub link. Work Experience

AI Engineer - Internship at GENERAL ERA DIGITAL SOLUTION JSC May 2022 - Sep 2022

– Developing a chatbot to support answering questions related to university admissions .

+ Gathered Vietnamese data related to university admissions, including questions and answers from websites, Facebook, Instagram, etc. Processed the collected data, removing noise such as icons, and then labeled the data for training the model.

+ Utilized GPT-2 and PhoBERT for training the model, followed by validation to determine the optimal model.

+ Employed RASA as a framework to deploy the optimized model and integrated it with Facebook for admissions inquiries.

– Techstack: Nature Language Processing, Deep Learning, PhoBert, RASA, Data mining, GPT-2, Data preprocessing. Skills

Technical Skills

– Programming Languages: Python, Java, C++, SQL.

– Machine Learning: Scikit-learn, TensorFlow, Keras, PyTorch.

– Data Analysis: Pandas, NumPy, Matplotlib, Seaborn.

– Frameworks : NLTK, SpaCy, Transformers (BERT, GPT-2,3), YOLO, GANs.

– Tools: Git, Docker, Jupyter Notebooks, Anaconda.

– English: Advanced

Soft Skills

– Problem-solving: Ability to analyze complex problems and develop effective solutions.

– Teamwork: Collaborative approach to working within multidisciplinary teams to achieve common goals.

– Time Management: Effective prioritization and organization of tasks to meet deadlines.

Contact this candidate