Qubestixx B.

Key Skills

Software

CVAT

Roboflow

Other

Top Subject Matter

Custom object detection (computer vision)

Sentiment analysis and entity extraction (NLP)

Image classification for African wildlife species (computer vision)

Top Data Types

Image

Text

Audio

Top Task Types

Object Detection

Entity (NER) Classification

Classification

Freelancer Overview

Computer Vision Annotation Portfolio (Personal Project). Core strengths include CVAT, Roboflow, and Google Sheets. AI-training focus includes data types such as Image and Text and labeling workflows including Object Detection, Entity (NER) Classification, and Classification.

Labeling Experience

Open-Source Data Contribution (Volunteer Work)

RoboflowImageClassification

Contributed 200+ validated image classifications to a public computer vision dataset focused on African wildlife species. Reviewed existing labels, identified 30+ mislabeled images, and reported issues to improve dataset hygiene. Maintained a consistent quality standard that was acknowledged by the dataset maintainer. • Performed classification-level validation on submitted image data. • Flagged and reported suspected labeling errors to the community dataset workflow. • Ensured annotations met dataset quality expectations prior to contribution. • Delivered contributions back to Roboflow Universe through the open-source process.

2026 - Present

NLP Text Annotation Exercise — Sentiment & Entity Tagging (Independent Learning)

OtherTextEntity (NER) Classification

Manually annotated a 3,000-row customer review dataset for sentiment polarity and key entity tags. Extracted product names and attributes using a spreadsheet-based tagging workflow. Defined guidelines to resolve ambiguous examples such as mixed sentiment and implied opinions. • Tagged sentiment as positive, neutral, or negative and performed entity extraction. • Used Python (Pandas) to analyze label distribution and address potential class imbalance. • Documented the end-to-end labeling process on GitHub from guidelines to quality analysis. • Simulated professional annotation lifecycle activities including labeling consistency checks.

2026 - Present

Computer Vision Annotation Portfolio (Personal Project)

CVATImageObject Detection

Built a custom object detection training dataset with 500+ labeled images. Applied bounding boxes and polygon masks across 10 object classes to support computer vision model training. Created and documented a labeling ontology to improve consistency for difficult cases. • Used CVAT and Roboflow for annotation and dataset management. • Exported datasets in COCO JSON, YOLO txt, and Pascal VOC XML formats. • Performed a 98% self-verified consistency check by re-labeling and comparing samples. • Documented edge cases such as occlusion, varied lighting, and background clutter.

2026 - Present

Education

Work History