Zhenyu Y.

Key Skills

Software

LabelImg

Labelbox

Top Subject Matter

Surface Defect Detection in Manufacturing

Automotive Connector Defect Detection

Top Data Types

Image

Top Task Types

Object Detection

Freelancer Overview

Researcher, Surface Defect Annotation for AI Training. Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal and Proprietary Tooling. Education includes Master of Science, 合肥大学 (2024) and Bachelor of Engineering, 合肥师范学院 (2021). AI-training focus includes data types such as Image and labeling workflows including Object Detection.

Labeling Experience

AI Data Labeler & Trainer (Automotive Detection)

ImageObject Detection

At Chery Automobile, I conducted data labeling and training work for defect detection in automotive manufacturing. My duties included labeling connector images with bounding boxes and classifications to train YOLOv8/v11-based AI models for object detection. I performed data annotation, quality checks, and collaborated on integrating labeled data with the detection pipeline. • Performed bounding box annotation on thousands of automotive connector images. • Created and validated training, validation, and test splits from the annotated image data. • Contributed to iterative model improvement based on evaluated labeling quality and detection outcomes. • Worked with cross-functional teams to deploy and assess models in production environments.

2024 - Present

Researcher, Surface Defect Annotation for AI Training

ImageObject Detection

I was responsible for collecting, annotating, and preparing image data to train computer vision models for surface defect detection. My role required implementing image annotation techniques, including adapting attention mechanisms and lightweight networks, to improve detection accuracy. I used algorithms such as SSD, Faster RCNN, and YOLO implemented in PyTorch to label and process the image data for model training. • Labeled and annotated product surface images for defect detection tasks. • Experimented with backbone networks and included attention mechanisms for annotation efficiency. • Improved YOLO-based models using labeled image datasets. • Reduced model size while maintaining high precision and recall rates.

2021 - 2024

Education

Work History