Susan Wambui

Key Skills

Software

Clickworker

CloudFactory

CrowdSource

CVAT

Remotasks

Sama

Telus

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Task Types

Bounding Box

Data Collection

Evaluation Rating

Point Key Point

Polygon

Freelancer Overview

I have extensive experience in data labeling and AI training data preparation, supporting the development of machine learning models across various domains including natural language processing, computer vision, and speech recognition. My work has involved curating high-quality datasets through meticulous annotation, classification, and entity recognition tasks, using tools such as CVAT, SRT tool, and custom annotation platforms. I’ve contributed to supervised learning workflows by designing annotation guidelines, ensuring inter-annotator agreement, and performing quality control on large-scale datasets, all while adhering to ethical data handling standards.

Labeling Experience

Image Segmentation

CVATImagePolygon

In a comprehensive CVAT project, I led the full-scale annotation of a complex image dataset, where the objective was to label every visible object in each frame — a process often referred to as "labeling everything." This required the use of polygon tools in CVAT to meticulously annotate diverse object classes, including people, vehicles, signage, street furniture, animals, and environmental elements like trees and buildings. The dataset included thousands of high-resolution images from urban and rural environments, demanding a high level of precision, consistency, and attention to edge cases.

In a comprehensive CVAT project, I led the full-scale annotation of a complex image dataset, where the objective was to label every visible object in each frame — a process often referred to as "labeling everything." This required the use of polygon tools in CVAT to meticulously annotate diverse object classes, including people, vehicles, signage, street furniture, animals, and environmental elements like trees and buildings. The dataset included thousands of high-resolution images from urban and rural environments, demanding a high level of precision, consistency, and attention to edge cases.

2024 - 2024

Education

Work History