Lê Nhân

Key Skills

Software

OpenCV AI Kit (OAK)

Roboflow

CVAT

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Text

Top Task Types

Bounding Box

Classification

Emotion Recognition

Mapping

Object Detection

Freelancer Overview

I have hands-on experience in data labeling and AI training data preparation, particularly for computer vision and satellite image classification tasks. My background in Artificial Intelligence has provided me with strong technical knowledge of deep learning, data preprocessing, and annotation workflows. I have worked on projects involving multispectral and RGB image datasets, ensuring high-quality labeled data for model training and validation. In addition to labeling accuracy, my knowledge of machine learning frameworks like PyTorch and TensorFlow allows me to align data labeling with downstream model requirements, ensuring efficient and meaningful AI training pipelines.

Labeling Experience

Multispectral Satellite Image Labeling for Land Cover Classification

RoboflowImageBounding BoxSegmentation

This project involved labeling and annotating multispectral Sentinel-2 satellite images from the EuroSAT and BigEarthNet datasets for land cover classification tasks. I was responsible for creating segmentation masks and assigning class labels (e.g., urban, vegetation, water, farmland) across multiple spectral bands. The project emphasized maintaining consistency in labeling standards, ensuring pixel-level accuracy, and verifying data quality through cross-validation and review cycles. The labeled data was later used to train and evaluate deep learning models (CNN and ViT) for explainable land cover mapping.

This project involved labeling and annotating multispectral Sentinel-2 satellite images from the EuroSAT and BigEarthNet datasets for land cover classification tasks. I was responsible for creating segmentation masks and assigning class labels (e.g., urban, vegetation, water, farmland) across multiple spectral bands. The project emphasized maintaining consistency in labeling standards, ensuring pixel-level accuracy, and verifying data quality through cross-validation and review cycles. The labeled data was later used to train and evaluate deep learning models (CNN and ViT) for explainable land cover mapping.

2025 - 2025

Traffic Sign Image Labeling for Object Detection

CVATImageBounding BoxClassification

This project focused on labeling and annotating traffic sign images to train AI models for automatic road sign detection. I created accurate bounding boxes around traffic signs and classified them by type (e.g., speed limit, warning, prohibition). The project required strict adherence to annotation guidelines to ensure high-quality labeled data for model training. The labeled dataset was used to train CNN and YOLO-based models to improve detection accuracy in real-world driving conditions.

This project focused on labeling and annotating traffic sign images to train AI models for automatic road sign detection. I created accurate bounding boxes around traffic signs and classified them by type (e.g., speed limit, warning, prohibition). The project required strict adherence to annotation guidelines to ensure high-quality labeled data for model training. The labeled dataset was used to train CNN and YOLO-based models to improve detection accuracy in real-world driving conditions.

2024 - 2024

Education

Work History