Ai Engineer Artificial Intelligence

CURRICULUM VITAE

PERSONAL INFORMATION

Full Name Trinh Anh Van

Nationality Vietnamese

E-mail *************@*****.***

SUMMARY

As an AI engineer, I specialize in developing cutting-edge solutions using artificial intelligence. My goal is to contribute to advancements in AI technology and solve real-world problems through innovative applications.

EDUCATION BACKGROUND

Oct 2024

- Bachelor’s Degree in Control Engineering and Automation of Hanoi University of Science and Technology (HUST)

Graduated: Excellence.

WORKING EXPERIENCE

Hanoi, Aimesoft

Sep 2024 - now

Computer Vision Engineer

- Research and develop vision products like OCR, pattern recognition, ...

- Develop backend system for AI serving.

TECHNICAL SKILLS

Main Programming

Languages

Python, C++, Matlab, Ladder…

Frameworks & Services Pytorch, Numpy, OpenCV, ezdxf, python-bioformat, vips-tools, … Other Skills Linux, Git, Docker, Control system, … PROJECTS

Lung Cancer Classification

- Extract WSI images from various file types like .vsi, .ets, .dtx,

.tmap.

- Survey some model for segmentation like CellViT, InstanceSeg, CelloType, …

- Processing data for training and infering with CellViT model.

- Training and finetuning CellViT model for nuclei segmentation.

- Optimize code for faster inference time.

- Extract special features for classification.

- Classify WSI image with those features.

- Packaged module with pyinstaller to a blackbox.

- Packaged code and deploy on sever.

Nishiokagu

- Developed a solution combining image processing, natural language processing (NLP), and other techniques to extract room and furniture information from AutoCAD .dxf files.

- Utilized the ezdxf library to efficiently extract information from

.dxf files..

- Implemented OpenCV's findContours function to identify room candidates based on geometric properties.

- Trained a DistilBERT model to classify text as either room- related or unrelated, enabling room detection through textual data.

- Integrated contours and text data to locate rooms and extract their coordinates.

- Applied contour detection to identify furniture candidates within floor plans.

- Trained a MobileNet-v2 model to classify contour images into specific furniture types.

- Packaged the solution into a deployable module and implemented it on a server for production use.

Ocr Translate

- Developed a project enabling users to upload files and select a language for translation.

- Using python-pptx to extract information from a .pptx file.

- Using python-docx to extract information from a .docx file.

- Using pandas to extract information from a .xlsx file or a .xls file.

- Using paddleOCR for extracting information from a pdf or an image file.

- Implemented translation functionality using the Google Translate API.

- Built an intuitive web interface using Gradio.

- Packaged the project as a module and deployed the application on a server for public access.

AI Digital

- Developed functionalities to assist visually impaired individuals.

- Designed a Pedestrian Light Detection Module to notify users when the pedestrian light is red (stop) or green (walk).

- Upgraded the model from SSD-MobileNetV2 to YOLO-v10 for enhanced accuracy in object detection.

- Optimized rules and algorithms to achieve higher system accuracy.

- Packaged and built the application using Inno Setup for streamlined deployment.

- Deployed the solution on an edge device (DE200) for real-time performance.

Wind turbine blade surface

classification

- Training a model for classify wind turbine blade surface.

- Using some basic model for classify like EfficientNet, ResNet, … for classify.

- Apply some Ensemble learning technique like Bagging, Boosting to get better performance.

- Apply some Few-shot learning technique like Siamese Network, Protocol Network for better performance.

LANGUAGE SKILLS

Vietnamese Native

English

- Good, can work in English environment

- Certificate for internal TOEIC 665

Contact this candidate