Crate Annotation for ML - Batch 13 - 10 Images
Budget: €30 – €250 EUR
This initial project is a small, well-defined batch to start our collaboration. We are looking for a partner for a long-term working relationship, and successful completion of this first task will lead to regular follow-up projects.
Project Scope
Task: Annotate a batch of 10 images.
Estimated Workload: We estimate this batch will take approximately 4-5 hours to complete.
Detailed Task Description
Your job is to precisely label the crates visible in the 10 provided images.
Annotation Type: You will be creating "Oriented Bounding Boxes" (rotated rectangles) that must accurately capture the orientation of the objects.
Classes: Each box must be assigned to one of two classes: small_crate or big_crate.
Software: The annotation will be done using a custom version of the labelImg software, which we will provide as a simple .exe file for Windows (no installation required).
Deliverables: The final annotations are automatically saved as text files (.txt).
Requirements
A meticulous, precise, and reliable work ethic.
A Windows operating system to run the provided .exe file.
Good communication skills (English or German).
Experience with image annotation (e.g., using labelImg, CVAT, etc.) is a plus but not mandatory, as a guide will be provided.
What We Provide
A folder containing the 10 images to be annotated.
A folder with examples of correctly completed annotations.
The necessary annotation software (labelImg.exe).
A detailed "Quick Guide" that explains the entire process and all important hotkeys.
Support for any technical questions or uncertainties.
Timeline: Till sunday
Project Scope
Task: Annotate a batch of 10 images.
Estimated Workload: We estimate this batch will take approximately 4-5 hours to complete.
Detailed Task Description
Your job is to precisely label the crates visible in the 10 provided images.
Annotation Type: You will be creating "Oriented Bounding Boxes" (rotated rectangles) that must accurately capture the orientation of the objects.
Classes: Each box must be assigned to one of two classes: small_crate or big_crate.
Software: The annotation will be done using a custom version of the labelImg software, which we will provide as a simple .exe file for Windows (no installation required).
Deliverables: The final annotations are automatically saved as text files (.txt).
Requirements
A meticulous, precise, and reliable work ethic.
A Windows operating system to run the provided .exe file.
Good communication skills (English or German).
Experience with image annotation (e.g., using labelImg, CVAT, etc.) is a plus but not mandatory, as a guide will be provided.
What We Provide
A folder containing the 10 images to be annotated.
A folder with examples of correctly completed annotations.
The necessary annotation software (labelImg.exe).
A detailed "Quick Guide" that explains the entire process and all important hotkeys.
Support for any technical questions or uncertainties.
Timeline: Till sunday