Bounding Box Labeler — Purple-Outlined Game Characters

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We connect people to hands-on projects that teach and shape how modern AI systems are built, and creating an OpenTrain account is free.

OpenTrain AI is the hiring and contracting organization for this project; you'll be contracted to complete image-labeling work through our platform and the project's labeling workspace.

About AI Training Work

AI training (also called data labeling or annotation) is the human work that teaches models to see and act: people annotate images, transcribe audio, rate model outputs, and more. This role focuses on image object-detection labeling to help train a YOLO model.

This kind of work is typically remote, flexible, and accessible — many projects require no formal degree, and contributors can build experience and move into higher-paying, specialized tasks over time.

• 100% remote: work from anywhere with a computer or phone and an internet connection.
• Flexible, part-time contracting work that fits around other commitments.
• Direct impact: your labels will become part of a YOLO training dataset used to improve model performance.

The Role

You will draw bounding boxes around characters that are outlined in purple in a set of game screenshots or images. Each image typically contains 1–5 objects to label.

Labels are created inside Roboflow and must be YOLO-compatible so they can be used directly for training. This is contract, part-time work: less than 20 hours per week, worldwide, paid per label at $0.02 each.

• Task: draw tight bounding boxes around purple-outlined characters in images.
• Tool: Roboflow labeling workspace; output must be YOLO-ready.
• Schedule: flexible, under 20 hours per week, contractor engagement.
• Compensation: $0.02 per labeled image (pay-per-label model).

What You'll Do

Follow the provided examples and labeling guidelines to mark every purple-outlined character with a precise bounding box. Each object seen in an image that meets the guidelines should be labeled.

Work efficiently while maintaining consistent, high-quality boxes so the dataset can be used for accurate object-detection training.

• Open each image in Roboflow and locate characters outlined in purple.
• Draw tight bounding boxes that include the full character but exclude excess background.
• Label every qualifying object in each image (1–5 objects per image is typical).
• Review your labels for overlap, missing objects, and consistency before submission.

Requirements

Intermediate experience labeling images for object-detection models intended for YOLO is required. You must be comfortable creating bounding boxes and producing YOLO-compatible annotations.

Familiarity with Roboflow or similar annotation tools is expected. You must accept per-label compensation of $0.02 and be available to work under 20 hours per week.

• Intermediate experience with object-detection labeling for YOLO.
• Experience using Roboflow or equivalent annotation software.
• Reliable internet access and a computer (or device supported by Roboflow).
• Ability to follow visual labeling guidelines and maintain consistency.

Who Should Apply

Apply if you have prior image-annotation experience for object detection, know how YOLO-format labels work, and can work accurately and efficiently in Roboflow. This project suits contributors building a portfolio of computer-vision labeling work.

Worldwide applicants are welcome; this is contract work with flexible hours, ideal for people looking for part-time labeling assignments.

• Annotators with YOLO/object-detection experience.
• Labelers familiar with Roboflow or similar tools.
• People seeking flexible, part-time contract work in AI training.

How It Works

Create or sign in to your OpenTrain account to apply. If selected, you'll receive access and instructions to the labeling workspace in Roboflow along with example images and guidelines to get started.

You will be paid per accepted label at the stated rate. Continue contributing as long as you meet quality standards and the project remains open.

• Apply through your OpenTrain profile and submit any requested experience details.
• Receive labeling access and examples, then complete tasks in Roboflow.
• Get paid per accepted label at $0.02 each; follow quality guidelines to ensure acceptance.