Obby Gee

Key Skills

Software

CVAT

Label Studio

Roboflow

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Image

Text

Video

Top Label Types

Action Recognition

Audio Recording

Bounding Box

Classification

Cuboid

Segmentation

Text Generation

Freelancer Overview

👋 Hi! I'm a data annotation specialist who helps AI and ML projects get the clean, accurate labeled data they actually need. Think of me as the person making sure your model doesn't confuse a cat with a stop sign 😊. I work with image and video data across the full range of annotation types: bounding boxes, polygons, semantic and instance segmentation, keypoints, cuboids, image masking, text and line annotation, and image classification. I follow project guidelines carefully and pay close attention to edge cases, because the small details are usually what separate a good model from a great one. I'm also a quick learner, so if you're using a tool I haven't worked with yet, I'll adapt fast and hit the ground running. Tools I work with: 🟣 CVAT 🟣 Label Studio 🟣 Roboflow 🟣 LabelMe 🟣 MakeSense.ai 🟣 VGG (VIA) 🟣 And whatever tool you might prefer in the future Output formats supported: 🟣 Pascal VOC 🟣 YOLO 🟣 JSON 🟣 COCO 🟣 CSV 🟣 Segmentation Mask If you need a reliable annotator who delivers consistent, well-structured datasets on time, let's talk.

Labeling Experience

Image Recognition Dataset for UI Component Detection

CVATImageBounding BoxObject Detection

Annotated a large-scale dataset of web interface screenshots to train a model for automatic UI component recognition. Applied bounding boxes to identify buttons, navigation bars, and input fields. Ensured strict adherence to pixel-perfect labeling standards to minimize noise in the model’s training phase. Also invested time specially focusing on labeling edge cases, such as responsive overlapping elements and dynamic pop-up modals, to ensure model robustness across different screen resolutions.

Annotated a large-scale dataset of web interface screenshots to train a model for automatic UI component recognition. Applied bounding boxes to identify buttons, navigation bars, and input fields. Ensured strict adherence to pixel-perfect labeling standards to minimize noise in the model’s training phase. Also invested time specially focusing on labeling edge cases, such as responsive overlapping elements and dynamic pop-up modals, to ensure model robustness across different screen resolutions.

2025 - 2025

Search Relevance and Query Categorization for E-commerce

Label StudioTextClassification

Assisted in the development of a search optimization engine. Tasked with evaluating the relevance of search queries against product results. Performed manual classification of over 5,000 data points to improve the precision of the search algorithm. Validated "Intent vs. Result" accuracy and flagged miscategorized items to refine the training dataset.

Assisted in the development of a search optimization engine. Tasked with evaluating the relevance of search queries against product results. Performed manual classification of over 5,000 data points to improve the precision of the search algorithm. Validated "Intent vs. Result" accuracy and flagged miscategorized items to refine the training dataset.

2025 - 2025

Education

Work History