Jacob Ngeke

Key Skills

Software

Appen

Clickworker

CloudFactory

Data Annotation Tech

Google Cloud Vertex AI

Labelbox

CVAT

Top Subject Matter

Autonomous vehicle systems

Traffic and road dynamics

simulation and testing

Top Data Types

Audio

Image

Text

Top Label Types

Data Collection

Evaluation Rating

Point Key Point

Polygon

Question Answering

Freelancer Overview

I have extensive experience in data labeling and AI training data management, with a strong focus on ensuring high-quality annotations for machine learning models. My expertise includes proficiency in various data annotation tools, such as Labelbox and Prodigy, which I have used to annotate diverse datasets, including images, text, and audio. I possess a keen eye for detail, which allows me to maintain accuracy and consistency in labeling, ensuring the data's integrity for successful model training. Throughout my career, I have contributed to multiple projects across different industries, including healthcare and autonomous vehicles, where I applied my domain knowledge to enhance the relevance of the training data. My background in programming, particularly in Python, has enabled me to automate labeling workflows and improve efficiency. I am also well-versed in quality assurance processes, ensuring that the labeled data meets the highest standards. My combination of technical skills, industry experience, and a commitment to data quality sets me apart as a strong candidate for any data labeling position.

Labeling Experience

Video Annotation & Data Labeling for AI Training

CVATVideoBounding BoxPolygon

Contributed to a large-scale AI training project by annotating and labeling video data to improve model accuracy. Tasks included frame-by-frame object detection, action recognition, scene classification and tracking. Followed strict labeling guidelines and quality assurance protocols, including inter-annotator agreement and review, ensuring high-quality, consistent datasets. The project involved handling 100+ hours of video data while maintaining >95% QA compliance.

Contributed to a large-scale AI training project by annotating and labeling video data to improve model accuracy. Tasks included frame-by-frame object detection, action recognition, scene classification and tracking. Followed strict labeling guidelines and quality assurance protocols, including inter-annotator agreement and review, ensuring high-quality, consistent datasets. The project involved handling 100+ hours of video data while maintaining >95% QA compliance.

2024 - 2024

Autonomous Vehicle Dataset Annotation for Object Detection

LabelboxImageBounding BoxPolygon

The project involved annotating a diverse dataset of over 50,000 images from urban and rural driving scenarios to create a comprehensive dataset for training object detection and segmentation models in autonomous vehicles. Specific tasks included annotating images with bounding boxes, creating segmentation masks for road areas and obstacles, and conducting instance segmentation to differentiate multiple instances of the same object. The project encompassed approximately 200,000 annotated objects across various categories. A multi-tiered quality control process was implemented, which involved regular audits to ensure guideline adherence. Additionally, inter-annotator agreement was measured to maintain consistency among annotators. Feedback loops were established to provide continuous improvement opportunities for the annotators' work.

The project involved annotating a diverse dataset of over 50,000 images from urban and rural driving scenarios to create a comprehensive dataset for training object detection and segmentation models in autonomous vehicles. Specific tasks included annotating images with bounding boxes, creating segmentation masks for road areas and obstacles, and conducting instance segmentation to differentiate multiple instances of the same object. The project encompassed approximately 200,000 annotated objects across various categories. A multi-tiered quality control process was implemented, which involved regular audits to ensure guideline adherence. Additionally, inter-annotator agreement was measured to maintain consistency among annotators. Feedback loops were established to provide continuous improvement opportunities for the annotators' work.

2023 - 2024

Education

Work History