Remote Sensing Vision-Language Model Fine-Tuning Data

RSCoVLM is a dataset for co-training vision-language models on remote sensing imagery for multi-task learning. The dataset is associated with a published academic paper and was created by a team of researchers including Qingyun Li, Shuran Ma, and Junwei Luo. It was last updated in January 2026.

Use Cases

• Fine-tune vision-language models on remote sensing imagery for multi-task learning as described in the associated paper.
• Apply co-training techniques to aerial imagery for improved model generalization in geospatial analysis.
• Use the dataset to benchmark multi-task learning performance on remote sensing tasks like image classification or captioning.

Strengths

• Dataset is linked to a peer-reviewed paper published in MDPI Remote Sensing.
• Dataset is associated with a detailed arXiv preprint (arXiv:2511.21272).
• The data is focused on the specific domain of remote sensing and aerial imagery.

Limitations

• No sample data, column definitions, or size information is available for inspection.
• The specific tasks, labels, and data structure are not described in the provided input.
• The dataset's creation method and source imagery are not detailed.

Provenance

Source

huggingface

Collection Method

Created for co-training vision-language models for remote sensing multi-task learning, as per the associated paper.

Freshness

Last updated on 2026-01-21.