Remote Sensing Vision-Language Foundation Models without Annotations via Ground Remote Alignment
This enables zero-shot, open-vocabulary tasks for satellite images, addressing a domain-specific need in remote sensing.
The paper tackles the problem of training vision-language models for remote-sensing images without textual annotations by using co-located internet imagery as an intermediary, resulting in a model that outperforms supervised methods with gains of up to 20% for classification and 80% for segmentation.
We introduce a method to train vision-language models for remote-sensing images without using any textual annotations. Our key insight is to use co-located internet imagery taken on the ground as an intermediary for connecting remote-sensing images and language. Specifically, we train an image encoder for remote sensing images to align with the image encoder of CLIP using a large amount of paired internet and satellite images. Our unsupervised approach enables the training of a first-of-its-kind large-scale vision language model (VLM) for remote sensing images at two different resolutions. We show that these VLMs enable zero-shot, open-vocabulary image classification, retrieval, segmentation and visual question answering for satellite images. On each of these tasks, our VLM trained without textual annotations outperforms existing VLMs trained with supervision, with gains of up to 20% for classification and 80% for segmentation.