Abstract: In recent years, the success of large-scale vision-language models (VLMs) such as CLIP has led to their increased usage in various computer vision tasks. These models enable zero-shot ...
Old-school C-Clamps are in every toolbox, but they have a lot more uses than just clamping things together while you work.
Abstract: This work explores capabilities of the pre-trained CLIP vision-language model to identify satellite images affected by clouds. Several approaches to using the model to perform cloud presence ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback