Metraipolis › archive › dc_b0f6fcdcbdb7

LongEarth-R1: Benchmarking and Aligning Vision-Language Models for Long-Horizon Earth Observation Reasoning — arXiv:2608.13344

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.13344
License (as archived)
CC0-1.0
Archived
2026-10-19T04:17:39.461Z
SHA-256 of the archived text
3306b76ab71b33b5ab88b271695fd4621259f7b85a40ac3ade6519aabe02b97c
Stable URL of this page
https://metraipolis.com/documento/dc_b0f6fcdcbdb7

How to cite

LongEarth-R1: Benchmarking and Aligning Vision-Language Models for Long-Horizon Earth Observation Reasoning — arXiv:2608.13344 — arXiv (en), https://arxiv.org/abs/2608.13344 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b0f6fcdcbdb7 (Metraipolis archive dc_b0f6fcdcbdb7, sha256 3306b76ab71b33b5ab88b271695fd4621259f7b85a40ac3ade6519aabe02b97c)
BibTeX
@misc{dcb0f6fcdcbdb7,
  title = {LongEarth-R1: Benchmarking and Aligning Vision-Language Models for Long-Horizon Earth Observation Reasoning — arXiv:2608.13344},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.13344},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b0f6fcdcbdb7; sha256 3306b76ab71b33b5ab88b271695fd4621259f7b85a40ac3ade6519aabe02b97c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.13344. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Long-horizon Earth observation reasoning requires models to organize multi-stage geographic evolution, localize spatial changes, detect temporal anomalies, and infer future from extended image sequences. However, existing remote sensing vision-language models mainly focus on isolated images, image pairs, or short sequences, limiting reliable grounding in the relevant frames and regions. We introduce LongEarth-Bench, a benchmark containing approximately 120k question-answering samples derived from 117k unique images. Its sequences average 15.14 frames and extend to 30 frames, covering 12 tasks across evolution summarization, spatial reasoning, anomaly identification, and logical prediction. A 30k-sample subset further provides structured reasoning traces linking key frames and changed regions to final answers. We develop LongEarth through supervised fine-tuning with explicit sequence identifiers and structured chain-of-thought supervision. Building on LongEarth, LongEarth-R1 applies group relative policy optimization with format, temporal, and spatial rewards. LongEarth-R1 achieves the best results on all 12 long-sequence tasks while remaining competitive on standard remote sensing benchmarks.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.13344 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T04:13:06.607Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).