Experienced AI engineer with a specialized focus on computer vision, I possess excellent algorithmic thinking and programming skills. Deeply passionate about AI, I am committed to continuous learning. Currently seeking opportunities to explore image processing to enhance proficiency in both technical and managerial aspects. My future goal is to establish and effectively manage a CV team. My career-long objective is to stay updated with evolving technological trends in computer vision.

Mining large datasets

Fine-tuning deep learning models, LLMs

Deploying models ONNX, TensorRT, Triton, Deepstream Algorithmic thinking and coding with C/C++, Python Quick learner with ability to work independently

Kafka, RabbitMQ

ElasticSearch

Docker

MySQL, SQLServer, MongoDB

Web programming with Python, HTML, CSS

ICOMM Media & Tech Jsc.

VisualSearch - Key Member

Develop Yolo, Retina detection model.

Research and deploy FastSAM (child of Segment-Anything) to segment all objects in images Develop core’s embedding model to extract characteristic features of objects Sep 2022 - Present

SUMMARY

THE HUNG PHAM

EXPERIENCE

SKILLS

AI Camera - Key Member

Apply several algorithms such as IVF, HNSW, Milnus for a search system Research and train face recognition module with large-scale face datasets Face Generative Module - Key Member

Face Restoration: GFP-GAN, GPEN, DifFace

Face Frontialization: Generate a frontialized face from a rotated face Face Attribute Editing: Research and build a module latent vector PTI for editing attribute’s face such as: smile, age.

Face Image Quality Assessment

Phone: +84-366****** Mail: ad3o3t@r.postjobfree.com Linkedin: linkedin.com/in/hungpham00 Github: github.com/ChuRuaNh0 Location: Ha Noi, Vietnam

Computer Vision Engineer

Other Modules - Member

License Plate Restoration & Car Restoration: Build restoration models for enhancing license plate images and car images quality to deal with practical problems. License Plate Detection: Research, train various detection methods (YOLO, Retinanet) to balance between accuracy and performance; deploy to Triton and Deepstream for production. Uniform Classification: Classify images of different uniforms into categories. Customize BatchNMS YOLO: Add BatchNMS to YOLO model to enable deployment on TensorRT OCR: Research, train various OCR methods: Combine various encoder and decoder pretrain English with data’s Vietnamese.

Diffusion Model (Image Generative)

Machine Learning

Neural Network and Deep learning

DeepLearning.AI - Stanford course on Coursera

DeepLearning.AI - Stanford course on Coursera

CERTIFICATES

AI Engineer - Data Science

Excellent AI Engineer - Data Science of VTC Academy FastSam (CASIA-IVA-Lab)

github.com/CASIA-IVA-Lab/FastSAM

Build and deploy engine onnx, tensorrt, NMS.

July 2023 - Present

Assistant Researcher in University

Hanoi university of Industry

A member of the CV team, researching ML algorithms for CV. Plant-disease-detection (3rd)

Suspicious-Human-Activity-Detection

Jun 2021 - Sept 2022

Faiss (Facebook)

https://github.com/facebookresearch/faiss

Optimize top-k search algorithm, range search algorithm with boundary search (on ram, on disk) Decrease complexity from O(nlog(n)) to O(nlog(k))

Sep 2023 - Present

CONTRIBUTIONS