Results for the few-and zero-shot prediction methods.

<p>The black line shows the performance of a ViT-14/L model initialised with CLIP pre-trained weights trained on the complete SON dataset. The line represents the same model, where only the linear probe is fine-tuned in a few-shot setting. The red line shows the performance of the ViT-14/L vis...

Full description

Saved in:

Bibliographic Details
Main Author:	Alex Levering (19762340) (author)
Other Authors:	Diego Marcos (19762343) (author), Nathan Jacobs (18263948) (author), Devis Tuia (1805620) (author)
Published:	2024
Subjects:	Medicine Science Policy Biological Sciences not elsewhere classified Information Systems not elsewhere classified fully supervised way enabled efficient transfer downstream tasks even based annotations highlights single linear layer shot methods combined landscape scenicness prediction trained vlm representation shot prediction potential shot methods shot learning model trained linear probe small number results show model fitted method outperforms language models image ratings image content e ., directly compared deep learning also explore aesthetic quality
Tags:	Add Tag No Tags, Be the first to tag this record!

Results for the few-and zero-shot prediction methods.

Similar Items