Oreofe Blessed

Key Skills

Software

Clickworker

Mercor

Mindrift

SuperAnnotate

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Task Types

Audio Recording

Classification

Data Collection

Evaluation Rating

Segmentation

Freelancer Overview

Results-driven Data Labeller with 2 years of experience in video data annotation and labeling, specializing in Atlas Capture projects. Proven track record of delivering high-quality labels for human action recognition tasks, adhering to dense labeling guidelines. Skilled in accurately identifying and labeling hand-object interactions, with a focus on precision and consistency. Strong understanding of AI training data requirements and best practices. Key skills include: video data annotation, human action recognition, dense labeling, object interaction labeling, and attention to detail. Proficient in using specialized labeling tools and software, with excellent communication skills for team collaboration via Discord. Experienced in working with video clips of humans completing physical tasks, with a focus on labeling meaningful actions and minimizing gaps in annotation.

Labeling Experience

Data Labeller

AppenVideoPoint Key PointSegmentation

The Atlas Capture Data Annotation & Labeling Program is a continuous initiative designed to ensure that all video data across the organization is accurately and comprehensively labeled in support of frontier AI research and model development. This program provides guidance for annotation teams to produce consistently correct, high-quality labels with maximum coverag referred to as Dense labeling. The program focuses on labeling video clips of humans completing a physical task. Dense labeling is the continuous, high-precision annotation of video data that captures every meaningful human action in the correct order of actions. Labels aim to fully represent observable behavior with minimal gaps, emphasizing accuracy and consistency. Only actions where the hand interacts with an object needs to be labelled. If a hand does not interact with an object, there’s no label required. I can then label the video clip with the “no action” button.

The Atlas Capture Data Annotation & Labeling Program is a continuous initiative designed to ensure that all video data across the organization is accurately and comprehensively labeled in support of frontier AI research and model development. This program provides guidance for annotation teams to produce consistently correct, high-quality labels with maximum coverag referred to as Dense labeling. The program focuses on labeling video clips of humans completing a physical task. Dense labeling is the continuous, high-precision annotation of video data that captures every meaningful human action in the correct order of actions. Labels aim to fully represent observable behavior with minimal gaps, emphasizing accuracy and consistency. Only actions where the hand interacts with an object needs to be labelled. If a hand does not interact with an object, there’s no label required. I can then label the video clip with the “no action” button.

2025

Education

Work History