Shaui F.

Key Skills

Software

Other

Top Subject Matter

Computer vision object detection dataset preparation and label QA

YOLO-based object detection dataset preparation and evaluation

AI output evaluation (QA/rubric-based assessment) for text and visual outputs

Top Data Types

Image

Text

Document

Top Task Types

Object Detection

Bounding Box

Polygon

Segmentation

Classification

Entity (NER) Classification

Point/Key Point

Polyline

RLHF

Text Summarization

Question Answering

Fine-tuning

Transcription

Evaluation/Rating

Text Generation

Freelancer Overview

Computer Vision Labeling and Review Project (Personal Project). Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Other. Education includes Bachelor’s Degree, Zhejiang University of Technology (2026). AI-training focus includes data types such as Image and Text and labeling workflows including Object Detection, Evaluation, and Rating.

Labeling Experience

AI Evaluation Practice (Independent Learning)

OtherText

Conducted ongoing AI output evaluation to judge the accuracy, completeness, logical rationality, and instruction compliance of AI-generated text and visual outputs. Identified common output issues such as logical contradictions, missing key information, hallucinations, and non-standard formatting. Applied consistent evaluation judgment criteria across batch text and image assessment tasks for high-precision, high-repeatability QA work.• Rated and compared AI outputs across multiple quality dimensions.• Flagged hallucinations, omissions, and formatting/non-compliance errors.• Used unified criteria to maintain consistent evaluation standards.• Performed batch review of large volumes of text and image outputs.

2023 - Present

YOLO Model Training Support (Personal/Academic Project)

OtherImageObject Detection

Supported YOLO model training by preparing pre-training datasets through image screening, deduplication, label sorting, and verification with standardized directory structure. Participated in model testing and evaluation by comparing batch predictions with manual ground truth labels and recording detection deviations. Optimized detection performance by tuning confidence and NMS thresholds and summarizing common failure scenarios linked to data and annotation quality.• Screened and deduplicated images for training readiness.• Verified and organized labels into YOLO-compatible dataset structure.• Evaluated model predictions against ground-truth labels and logged deviations.• Tuned inference thresholds and compiled error/failure analyses for iterative data improvement.

2024 - 2024

Computer Vision Labeling and Review Project (Personal Project)

OtherImageObject Detection

Annotated large-scale image datasets for object detection using bounding box marking and category classification labels across daily and industrial scenes. Performed secondary inspection to calibrate and correct labeling inconsistencies, including category mislabels, offset bounding boxes, and missing annotations. Analyzed model detection defects by tracing false positives, missed targets, and confusing edge cases to drive standardized data quality rules.• Labeled and reviewed bounding box/category annotations for multi-scene datasets.• Conducted quality inspection and corrected labeling errors to unify standards.• Compared inference defect types (false positives/missed targets/edge confusion) against label issues.• Documented practical data-quality rules to improve model training generalization.

2024 - 2024

Education

Work History