I have created a public project on object classification. Project is created to classify smartfarm baskets. I have several annotated images and I plan to annotate the remaining unannotated data using ‘Autodistill’ library. But There is no way I can download this unannotated image set. I have tried several methods and I have searched every where in google. There was a documentary saying how to download unannotated data but it doesn’t have clearly mentioned how to do that. Please help me with this.
To download your unannotated images from Roboflow, you can use the roboflow-python package. Here is an example of how to do it:
import roboflow
rf = roboflow.Roboflow()
project = rf.workspace(workspace_id).project(project_id)
# download a new version
version_num = project.generate_version()
version = project.version(version_num)
version.export("coco-segmentation")
version.download(model_format="coco-segmentation", location=output_dir)
In this code, replace workspace_id and project_id with your workspace and project IDs respectively. The location parameter in the version.download() function should be the directory where you want to save your images.
To download unlabelled images, you can use the API behind Roboflow Search. By default, the notebook loads 500 unlabelled images by performing a search in a loop (loading the maximum of 250 at a time).
At the end of this process, you will have a directory unlabelled/ with your unlabeled images.
For more details, you can refer to the blog post you mentioned.
Once you have your unannotated images, you can use the Autodistill library to automatically label them. The Autodistill library can train a model based on folder names with images in them. This can be particularly useful if your images are organized in folders by class. For more information on how to use Autodistill, you can refer to this blog post.
Hi! I’m Roboflow’s AI powered bot. This is an automated response powered using AI with knowledge from our docs, blog, knowledge base, and previous forum responses. If you find this to be a helpful answer, please mark it as the solution using the checkbox icon below. If this doesn’t solve your issue, please follow up with what you’ve tried/why, and the community will continue to chime in as usual.