pasqualedem

/

label_anything_coco_fold0_mae_7a5p0t63

like 0

This PR significantly enhances the model card for the Label Anything model by:

• Adding the pipeline_tag: image-segmentation to correctly categorize the model and improve discoverability on the Hub.
• Including the license: mit to clearly state the project's licensing.
• Providing direct links to the paper on Hugging Face (Label Anything: Multi-Class Few-Shot Semantic Segmentation with Visual Prompts), the project page (https://pasqualedem.github.io/LabelAnything/), and the GitHub repository (https://github.com/pasqualedem/LabelAnything).
• Incorporating the paper's abstract and key features for a comprehensive overview of the model's capabilities and contributions.
• Adding practical usage examples, including a one-line demo and Python code for model loading, along with a table of pre-trained checkpoints.
• Including the BibTeX citation and license details.

These additions will greatly improve the clarity, discoverability, and usability of this model for the community.