How to Use the External Labelling Service on Datature Nexus
Related AI
In this video
Enhancing Data Annotation with External Labeling Services on Datature Nexus Platform
The rapid growth of artificial intelligence (AI) and machine learning (ML) applications has brought with it a surge in the demand for high-quality labeled data. Datature Nexus platform addresses this need by providing a robust and flexible platform for annotation and labeling. However, for some businesses, the volume of data to be annotated may exceed their current capacity or they might be faced with a strict deadline. This is where Datature's External Labeling Services come in, offering a straightforward and secure way to accelerate the annotation process.
The Need for External Labelers
There are several reasons why a company might consider leveraging external labeling services. These include:
- Having an extensive amount of visual data that needs to be annotated.
- A lack of sufficient team members or collaborators to perform the annotation tasks.
- A strict deadline that requires a swift completion of data annotation.
Requesting External Labeling Services on Datature Nexus
The process of requesting external labeling services on the Datature Nexus platform is straightforward and intuitive. From the home page of your project, you can navigate to the «Hire Labelers» button, which will guide you through a step-by-step process to submit your request.
Step 1: Project Description and Budget
In the first step, you provide a brief description of your project, detailing the specifics of what you want to label. For example, in an animal detection project, you might specify that you need to label dolphins and flamingos. You also provide your annotation budget, which will determine the number of images that can be annotated.
Step 2: Sample Annotations
The next step involves providing sample annotations. This is a crucial part of the process as the sample annotations act as the ground truth and golden standard for the external labelers. These should include representative images common in your dataset, along with some edge cases that may require more nuanced industry knowledge for accurate annotation.
Step 3: Cost Estimation and Plan Selection
Based on your budget and the complexity of the task, you can choose from three different plans:
- The Basic Plan: Offers up to two curated labelers to complete your annotations.
- The Standard Plan: Provides a dedicated team without a strict deadline.
- The Priority Plan: Comes with a dedicated team and a strict deadline that you can set according to your preference.
Step 4: Number of Images to Annotate
You can specify the number of images you would like to have annotated, which does not necessarily have to be the total number of unlabeled images in your dataset.
H4: Step 5: Reviewing Your Request
The final step is to review the details you have inputted in the previous steps, including your project description, the number of images to annotate, and your chosen plan. Before submitting your request, you should also review the terms and conditions.
Tracking Your Request and Ensuring Data Security
Once your request has been submitted, you can view the scheduled job on the «Scheduled Jobs» page. Here, you can track the progress of your request, including the number of images annotated, in review, to be fixed, completed, and not started.
Throughout the entire process, data security is of utmost importance. The annotations will be done on the platform, ensuring that no data leaves it at any point. External annotators have temporary accounts and can only access the data requested for annotation, with any other private or sensitive data being restricted from their view.
To sum up, the external labeling services on the Datature Nexus platform offer a convenient and secure solution to scale up your data annotation efforts, ensuring you meet your project deadlines without
Tags
Datature