Creating a Project
How do I Create a Project?
-
Go to the Workspace Dashboard - The Workspace Dashboard is the first page the user lands on immediately after logging in or creating an account. You can also access the dashboard by clicking the Nexus logo on the top left corner.
-
Click "Create New Project ⊕" - The button is located right below the Projects Dashboard header.
-
Select Project Type - This sets
Type Description Classification Your goal will be to categorize images into predefined classes, learn visual patterns and features associated with each class, and apply consistent labelling to new and unseen data. Object Detection Your goal will be to identify and classify each object of interest and predict bounding boxes around detected objects. Instance Segmentation Your goal will be to detect and classify distinct object instances and outline each object with polygons. Segmentation Your goal will be to assign each pixel in the image to predefined classes, differentiate between overlapping and adjacent of objects, and understand full-scene context with bounding boxes and polygons. Keypoint Your goal will be to identify individual objects as well as identify and label key anatomical or structural points. Bitmask Your goal will be to define each pixel for individual objects and analyze all objects as binary bit mask representations. Medical 3D Your goal will be to analyze volumetric medical imaging data to detect 3D structures through voxel-level masks. Medical Classification Your goal will be to analyze medical imaging data to learn to classify entire medical images to predefined classes. Support for other types and content will be released in future updates.
-
Select Data Type - This sets
Type Description Compatible Project Types Image Use static image files as the foundation for your AI workflows. This data type supports a wide range of visual tasks where spatial and appearance-based features are key. By selecting image data, your pipeline can be optimized to:
- Analyze individual image frames independently of time or sequence
- Support annotations such as classification, bounding boxes, segmentation, and keypoints
- Adapt easily to diverse domains from natural images to scanned documents and medical scansObject Detection, Classification, Segmentation, Instance Segmentation, Keypoint, Bitmask Video Use video sequences to capture temporal dynamics and contextual information across frames. This data type is ideal for time-sensitive tasks and moving object analysis. By selecting video data, your pipeline can:
- Understand motion patterns and temporal context over time
- Track objects persistently across multiple frames
- Annotate frame-by-frame sequences, transitions, and behaviorsObject Detection, Classification, Segmentation, Instance Segmentation, Keypoint, Bitmask Image & Video Combine the strengths of static image analysis with the temporal insights of video data. This hybrid data type supports use cases that require both spatial precision and motion understanding. By selecting image and video data, your pipeline can:
- Analyze both still images and video sequences in a unified workflow
- Support classification, bounding boxes, segmentation, and object tracking
- Leverage spatial features and temporal cues for improved accuracyObject Detection, Classification, Segmentation, Instance Segmentation, Keypoint, Bitmask NIfTI & DICOM Use standard medical imaging formats like NIfTI and DICOM to enable 3D volumetric analysis across slices. This data type is optimized for clinical and research workflows involving CT, MRI, PET, and other scan modalities. By selecting NIfTI or DICOM data, your pipeline can:
- Interpret volumetric scan data with spatial and intensity consistency
- Perform slice-by-slice or full-volume segmentation and labeling
- Integrate metadata, such as modality or patient orientation, into the AI workflowMedical 3D, Medical Classification -
Key in your Project Name
-
Key in your Project Description
-
Select Project Localization
To ensure dedicated project residency in a specific region, select your preferred location during project creation. Selecting certain regions predetermines data residency in the same region.
Region Target User Residence Multi-Region Anywhere Europe [EU-Belgium] European Union North America [US-CENTRAL] North America United Kingdom [UK-LONDON] United Kingdom Asia [ASIA-SOUTHEAST] Asia -
Invite team members for collaboration
-
-
Click "Next step" - Your new project is now created! 🎉
Ensure that you are in the correct workspace before creating your project. To switch to a different workspace, click on Navigating Between Workspaces.
Once you have created your project, you can now use all the project management functionality that we have available. To see what else you can do with your project, go to Managing Your Projects. To manage collaborators on your project, go to Managing Your Project Collaborators. To start importing your project data, go to Asset Upload and Annotation Upload.
👋 Need help? Contact us via website or email
🚀 Join our Slack Community
💻 For more resources: Blog | GitHub | Tutorial Page
🛠️ Need Technical Assistance? Connect with Datature Experts or chat with us via the chat button below 👇
Updated 7 days ago