Keith Bradley

Key Skills

Software

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Image

Video

Top Label Types

Bounding Box

Point Key Point

Segmentation

Classification

Tracking

Emotion Recognition

Freelancer Overview

I am an AI Training Specialist with 3 years of hands-on experience in data labeling and annotation for image, video, and audio datasets. My work has focused on supporting machine learning and computer vision projects by creating high-quality datasets using tools like Labelbox, YOLO, CVAT, Supervisely, and VGG Image Annotator. I am skilled in object detection, segmentation, keypoint annotation, and object tracking, and have a strong track record of maintaining accuracy and consistency across large-scale annotation projects. I have collaborated closely with machine learning engineers to refine training data for computer vision and NLP models, and I am experienced in quality assurance, data validation, and efficient task management to meet demanding project deadlines. My attention to detail and commitment to data quality help ensure optimal model performance and reliable AI solutions.

Labeling Experience

Advanced Video Object Detection & Multi-Object Tracking Project

LabelboxVideoBounding BoxPoint Key Point

Currently contributing to a high-volume video annotation project focused on improving real-time object detection and multi-object tracking models. The project involves frame-by-frame annotation of complex urban and dynamic scenes to enhance model accuracy in motion-based environments. Key responsibilities include: Annotating bounding boxes across consecutive video frames Performing multi-object tracking with consistent ID assignment Segmenting moving objects in crowded environments Labeling action recognition sequences (walking, running, vehicle turning, etc.) Annotating keypoints for pose estimation and motion analysis Handling occlusion, motion blur, low-light, and fast-movement scenarios The project consists of thousands of video sequences, each requiring strict adherence to annotation guidelines. I maintain high temporal consistency and quality control standards, achieving strong reviewer agreement scores and minimal correction rates.

Currently contributing to a high-volume video annotation project focused on improving real-time object detection and multi-object tracking models. The project involves frame-by-frame annotation of complex urban and dynamic scenes to enhance model accuracy in motion-based environments. Key responsibilities include: Annotating bounding boxes across consecutive video frames Performing multi-object tracking with consistent ID assignment Segmenting moving objects in crowded environments Labeling action recognition sequences (walking, running, vehicle turning, etc.) Annotating keypoints for pose estimation and motion analysis Handling occlusion, motion blur, low-light, and fast-movement scenarios The project consists of thousands of video sequences, each requiring strict adherence to annotation guidelines. I maintain high temporal consistency and quality control standards, achieving strong reviewer agreement scores and minimal correction rates.

2023

Autonomous Vehicle Object Detection & Tracking Annotation Project

LabelboxImageBounding BoxPoint Key Point

Led large-scale image and video annotation for an autonomous vehicle computer vision project focused on real-time object detection and multi-object tracking. The project involved annotating over 250,000+ images and 3,000+ video sequences captured in urban and highway environments. Key responsibilities included: Creating precise bounding boxes for vehicles, pedestrians, cyclists, traffic signs, and road infrastructure Performing semantic and instance segmentation for road scenes Annotating keypoints for pedestrian pose estimation Conducting frame-by-frame object tracking across video sequences Performing quality assurance reviews to maintain annotation consistency and accuracy Validating dataset integrity before model training Maintained a 98%+ annotation accuracy rate by following strict labeling guidelines and multi-level QA processes. Collaborated closely with ML engineers to refine labeling schemas and improve model performance metrics such as mAP (mean Average Precision).

Led large-scale image and video annotation for an autonomous vehicle computer vision project focused on real-time object detection and multi-object tracking. The project involved annotating over 250,000+ images and 3,000+ video sequences captured in urban and highway environments. Key responsibilities included: Creating precise bounding boxes for vehicles, pedestrians, cyclists, traffic signs, and road infrastructure Performing semantic and instance segmentation for road scenes Annotating keypoints for pedestrian pose estimation Conducting frame-by-frame object tracking across video sequences Performing quality assurance reviews to maintain annotation consistency and accuracy Validating dataset integrity before model training Maintained a 98%+ annotation accuracy rate by following strict labeling guidelines and multi-level QA processes. Collaborated closely with ML engineers to refine labeling schemas and improve model performance metrics such as mAP (mean Average Precision).

2021 - 2023

Education

Work History