Hi! I am ThorPham,
I started coding when I was 25. My major is banking and finance but I am passionate about IT, I taught myself everything related to programming. I love computer vision, image processing .
Tool and framework
- Programming Language : Python, JavaScript
- Web development : Reactjs, Flask, FastAPI
- Database : Mongo, Firebase
- Devops : Git, Docker, Bash script Shell
- Deep learning Framework : Pytorch, Tensorflow
- Image processing : Numpy, OpenCV, Pillow
- Machine Learning : Scikit-learn
Interest Fields
- Mathematics : Graph Theory, Graph Algorithm, Linear Algebra, Calculus, Optimization
- Computer Vision : Image Processing, Object Detection, Image Segmentation, OCR
- Image Processing : Traditional Image Processing Algorithm
- Computer science : Data Structure and Algorithm, Computer Graphics, Computer Architecture
- Natural language processing : Sequence to sequence, Word embbeding, Transfomer
- Statistic : Bayesian statistics, Graphical Models.
Publications
Deep Learning Approach for Receipt Recognition (2019)
Anh Duc Le, Dung Van Pham, Tuan Anh Nguyen
License Plate Recognition base on Deep Neural Network (2019)
Anh Duc Le, Dung Van Pham
Facial Landmark Detection with Learnable Connectivity Graph Convolutional Network (2022)
Le Quan Nguyen , Dung Van Pham, Yanfen Li , Hanxiang Wang , Lien Minh Dang, Hyoung-Kyu Song , Hyenjoon Moon
Key Information Extraction from Mobile-Captured Vietnamese Receipt Images using Graph Neural Networks Approach (2022)
Dung Van Pham, Le Quan Nguyen, Nhat Truong Pham, Bao Hung Nguyen, Duc Ngoc Minh Dang, Sy Dzung Nguyen
Vietnamese Scene Text Detection and Recognition using Deep Learning: An Empirical Study (2022)
Nhat Truong Pham, Dung Van Pham, Qui Nguyen Van, Bao Hung Nguyen, Duc Ngoc Minh Dang, Sy Dzung Nguyen
Personal Projects
Simple OCR
- Feature and Information: Detect and recognition for vietnamese and english language. It extracts text accurately even in images with severe distortion or complexity. Model OCR trained on 5M dataset with 1M real and 4M synthetic.
- FrontEnd : Reactjs, Pure Css, Canvas
- UX/UI : Clone and follow by Clova AI
- Backend : Pytorch, FastAPI
- Web Demo :demo
- Video Demo :demo
- Source code :code
Remove Background
- Feature and Information : Automatic remove background. It works 100% automatically: You don't have to manually select the background/foreground layers to separate them. You can login with google and save data in firebase store
- FrontEnd : Reactjs, Pure Css
- UX/UI : Clone and follow by remove.bg
- Backend : Pytorch, FastAPI
- Database : FireBase
- Web Demo :demo
- Video Demo :demo
- Source code :code
ALPR
- Feature and Information : Automatic License Plate Recognition system that works in all environments, optimized for your location. Can work with plate vietnamese with 1 line or 2 line and run only CPU with hight accuracy.
- FrontEnd : Reactjs, Pure Css
- UX/UI : Clone and follow by platerecognizer.com
- Backend : Pytorch, FastAPI
- Web Demo : demo
- Video Demo :demo
- Source code :code
Invoice OCR
- Feature and Information : Automating Receipt Digitization with OCR . The system can extract information from receipt image, with customizable fields, and extensible with national identity card , business cards or financial statements.
- FrontEnd : Reactjs, pure Css
- UX/UI : Clone and follow by nanonets.com
- Backend : Pytorch, FastAPI
- Web Demo : demo
- Video Demo :demo
- Source code : code
Certificates
- Modern JavaScript: ES6 Basics Coursera Link
- Object Oriented Programming in JavaScript Coursera Link
- Linear Regression with NumPy and Python Coursera Link
- Principal Component Analysis with NumPy Coursera Link
- Image Compression with K-Means Clustering Coursera Link
- Named Entity Recognition using LSTMs with Keras Coursera Link
- Neural Network Visualizer Web App with Python Coursera Link
- Programming Fundamentals Coursera Link
- Project: Creating Your First C++ Application Coursera Link
- Capstone Project GUVI Geek Networks, IITM Research Park Geek Networks Link
- Feedforward Neural Networks GUVI Geek Networks, IITM Research Park Link
- GIT GUVI Geek Networks, IITM Research Park Link
- Game development using PyGame GUVI Geek Networks, IITM Research Park Link
- Introduction to Pytorch GUVI Geek Networks, IITM Research Park Link
- JavaScript GUVI Geek Networks, IITM Research Park Link
- Primitive Neurons GUVI Geek Networks, IITM Research Park Link
- Sigmoid Neuron GUVI Geek Networks, IITM Research Park Link
- Training Feedforward Neural Networks GUVI Geek Networks, IITM Research Park Link
- Convolutional Neural Networks GUVI Geek Networks, IITM Research Park Link
- Encoder Decoder Models GUVI Geek Networks, IITM Research Park Link
- Introduction to Object Detection GUVI Geek Networks, IITM Research Park Link
- Deep Convolutional Neural Networks GUVI Geek Networks, IITM Research Park Link
- Optimization Algorithms GUVI Geek Networks, IITM Research Park Link
- Sequence Models GUVI Geek Networks, IITM Research Park Link
- Introduction to TensorFlow for Artificial Intelligence, Machine Learning, and Deep Learning Coursera Link
- Convolutional Neural Networks Coursera Link
- Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization Coursera Link
- Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization Coursera Link
- Mathematics for Machine Learning: Linear Algebra Coursera Link
- Matrix Methods Coursera Link
Awards and Honors
Top 5 ZALO hackathon (2019)
Topic : Illegal Nine-dash Line Detector
Incentive award in AI4VN challenge (2020)
Topic : Smart Citizens for Smart Cities
Incentive award in AI challenge HCM (2020)[Link]
Topic : Multi-Class Multi-Movement Vehicle Counting
Incentive award in AI challenge HCM (2021)[Link]
Topic : Vietnamese Scene Text Recognition