Xin L.

Key Skills

Software

Appen

CloudFactory

Top Subject Matter

Computer Vision

Natural Language Processing/Sentiment Analysis

AIGC Educational Content Generation

Top Data Types

Image

Text

Top Task Types

Bounding Box

Classification

Segmentation

Freelancer Overview

Data Annotation Lead for Object Detection Model Training. Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include OpenCV, Internal, and Proprietary Tooling. Education includes Bachelor of Science, 安徽理工大学 (2024). AI-training focus includes data types such as Image and Text and labeling workflows including Bounding Box, Classification, and Segmentation.

Labeling Experience

Image Data Labeling and Segmentation for AI Content Generation

ImageSegmentation

Directed OCR recognition and adaptation of AI learning content generation tool using large model APIs and LoRA fine-tuning. Labeled and categorized image data of handwritten notes, formulas, and diagrams for knowledge point extraction and concept mapping. Supervised creation of annotated datasets for automated summarization, mind map generation, and formula transcription. • Focused on image-level annotation for educational structure understanding. • Integrated AI labeling in workflow to improve knowledge recognition accuracy. • Refined dataset for improved model performance in real-world educational scenarios. • Led labeling technique improvements tailored to professional terminology and formulas.

2026 - 2026

Text Data Labeling and Annotation for Sentiment Analysis

TextClassification

Built a high-quality labeled text dataset for a Transformer-based sentiment classification AI model. Conducted web scraping, text deduplication, stop words filtering, and tokenization to structure and annotate the data for multi-class sentiment tasks. Guided the preprocessing and labeling pipeline to ensure accuracy and applicability for real-world sentiment analysis and feedback filtering. • Included hands-on annotation of positive, negative, and neutral sentiment. • Designed processes to create a robust training dataset for NLP. • Used techniques to enhance data sample representativeness and model reliability. • Facilitated model fine-tuning with clean, labeled text data.

2025 - 2025

Data Annotation Lead for Object Detection Model Training

ImageBounding Box

Led data selection, cleaning, annotation, and augmentation for YOLOv8-based smart object detection system. Applied random cropping, flipping, and color space transformation to enrich the image dataset and mitigate model overfitting. Managed dataset preparation to enhance object recognition, focusing on optimizing training samples for multi-class real-time detection. Oversaw real data labeling workflow, supporting research on detection efficiency and minimizing missing small object cases. • Tasks included selection, cleaning, and annotating of object detection datasets. • Implemented data augmentation to diversify training images. • Ensured dataset quality for model accuracy improvements. • Used labeling to drive targeted model parameter tuning and validation.

2025 - 2025

Education

Work History