A project houses the images and annotations for your project. All projects exist within one workspace. Once images are annotated, they can generate a version, which is a copy of your images and annotations at the time, augmented and processed with the configuration you select.
A project has a name, a type (ex: object detection) and an annotation group. For more information, take a look at the process of creating a project.
Create a Project
First, go to the Roboflow dashboard. Then, click "Create New Project":
You will be asked to specify:
A project type. Here is a brief summary of each project type:
Object Detection: Find the location of objects in an image.
Single-Label Classification: Given a limited set of categories, assign a label to an image.
Multi-Label Classification: Given a limited set of categories, assign an arbitrary number of labels that are relevant to the image.
Instance Segmentation: To the pixel level, find the location of objects in an image.
Semantic Segmentation: To the pixel level, find the location of objects in an image and create unique references for each object found.
Keypoint Detection: Find the location of objects and their keypoints in an image. Commonly used for determining the pose of an object.
A project name: The name of your project.
What you are detecting: A label that summarizes what you are detecting.
When you have specified these values, submit the form to create the project.
If you would like to see another type of project supported you can select the option from the dropdown of project types to indicate your interest.