Pith. sign in

REVIEW 3 major objections 5 minor 32 references

FusionRS: A Large-Scale RGB-Infrared-Style Remote Sensing Dataset for Cross-Modal Vision-Language Learning

T0 review · 3 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A new 600,000-pair dataset of RGB–infrared–text triplets, built by translating RGB remote sensing images into infrared style, substantially improves infrared-to-text retrieval and infrared-aware captioning in vision-language models.

desk verdict A genuinely useful synthetic RGB–IR–text dataset asset, but the central claim of transfer to real sensor-captured IR is untested and the abstract overstates it. read the letter →

arxiv 2606.17020 v2 pith:IIAFWVPS submitted 2026-06-15 cs.CV cs.AI

Jiaju Han , Ben Zhang , Xuemeng Sun , Qike Zhang , Yuxian Dong , Dingyi Lu , Chengyin Hu , Luwei Yang
show 3 more authors
Fengyu Zhang Yiwei Wei Jiujiang Guo
This is my paper · ORCID
classification cs.CVcs.AI
keywords RGB-infrared-textdatasetvision-languagelearningremotesensingcross-modalretrievalcontrastiveinfrared-awarecaptioningsyntheticinfrareddiffusion-basedtranslation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that the main obstacle to infrared-aware remote sensing vision-language models is the lack of large-scale RGB–infrared–text data, and that this gap can be filled synthetically. It introduces FusionRS, which pairs 600,000 RGB remote sensing images with diffusion-translated infrared-style counterparts and with captions — 45,913 of them rewritten to describe infrared cues such as intensity, contrast, and structure. Training CLIP-style and generative vision-language models on FusionRS improves RGB–IR alignment, infrared-to-text retrieval, and infrared-aware captioning and question answering compared with RGB-only and non-IR-aware settings. The authors argue that IR-aware textual supervision is the crucial ingredient for grounding language in infrared patterns, and that their pipeline offers a scalable route to infrared-capable remote sensing VLMs without paired sensor data.

What carries the argument

The engine of the paper is the RGB–IR–text triplet combined with a tri-modal contrastive objective, L_CLIP = 1/3(L_r,c + L_v,c + L_r,v), where L_r,c aligns RGB with text, L_v,c aligns infrared with text, and L_r,v explicitly aligns the two visual modalities. The aligned RGB–IR pairs are created by DiffV2IR, a diffusion-based visible-to-infrared translator, and the IR-aware captions are generated by a large multimodal language model from the RGB image, the IR image, and the original caption. This triplet structure lets the model learn both cross-modal visual correspondence and infrared–language grounding in one shared embedding space, and it is the component that the ablations show to be esse

What would settle it

Evaluate a FusionRS-trained CLIP model and VLM on real RGB-IR remote sensing pairs (e.g., VEDAI, DroneVehicle) for IR-to-text retrieval and IR-cue QA. If performance on real IR drops substantially relative to synthetic IR, or fails to beat an RGB-only baseline, the central claim of transferability is falsified. A quicker check: measure the distribution gap between synthetic IR embeddings and real IR embeddings of the same scenes; if the gap is large, or if RGB-IR alignment is trivially high because the synthetic IR is nearly a grayscale copy, the modality gap is not genuine.

Watch

Extended reading notes

Core claim

FusionRS is the first large-scale RGB–infrared–text dataset for general remote sensing vision-language learning, built by translating 600,000 public RGB remote sensing images into infrared-style counterparts with a diffusion-based translator. Each triplet keeps the original scene caption, and a curated subset adds 45,913 IR-aware captions that explicitly describe intensity, contrast, texture, and structure while preserving scene semantics. Training CLIP-style models with a tri-modal contrastive objective (RGB–text, IR–text, RGB–IR) and then fine-tuning generative VLMs with IR-aware captions yields consistent improvements in IR-to-text retrieval, RGB-IR retrieval, and infrared-aware captionin

Load-bearing premise

The load-bearing premise is that DiffV2IR's synthetic infrared-style images are a faithful enough proxy for real infrared observations that models trained on them will transfer to sensor-captured infrared data; the paper does not test any real RGB-IR pairs.

Editorial extensions

If this is right

  • FusionRS enables training of infrared-capable remote sensing VLMs from RGB-only archives, removing the need for physically paired infrared sensors.
  • Models trained on the dataset show consistent gains in IR-to-text retrieval and RGB-IR alignment across multiple CLIP backbones, suggesting the dataset transfers across architectures.
  • IR-aware captions are critical: fine-tuning on them (alone or mixed with original captions) lifts IR-cue QA and caption quality dramatically over original-caption-only training.
  • The dataset supports both contrastive representation learning and generative instruction tuning, so it can serve as a foundation and benchmark for future RGB-IR vision-language research.
  • Held-out test splits for retrieval, captioning, and VQA provide a standardized evaluation protocol for the community.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If synthetic-to-real transfer holds, the same pipeline could be applied to other non-RGB modalities (e.g., SAR, thermal, hyperspectral) to create large cross-modal VLMs from existing RGB archives, a direction the paper does not pursue.
  • The consistent superiority of the combined supervision setting (original + IR-aware + RGB-IR alignment) suggests a general recipe for cross-modal dataset design: pair raw captions with modality-specific captions and explicit inter-modal alignment, rather than relying on one supervision type.
  • The paper's evaluation is entirely on synthetic IR, so the decisive test is whether FusionRS-trained models transfer to real sensor-captured infrared data (e.g., VEDAI, DroneVehicle); until that test is run, the claimed gains are demonstrated only on the synthetic distribution.
  • Generating IR-aware captions with a 72B model is expensive; a cheaper distilled captioner or rule-based templates might reproduce the gains, but the paper does not ablate this.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper introduces FusionRS, a 600,000-pair RGB–infrared–text dataset for remote sensing, constructed by translating RGB images from five public datasets into synthetic infrared-style images with DiffV2IR and by generating IR-aware captions for a 49K-image subset with Qwen2.5-VL-72B. The authors train CLIP-style models with a tri-modal contrastive objective and fine-tune generative VLMs with original, IR-aware, or mixed captions. They report improved IR-to-text retrieval and IR-aware captioning/VQA over RGB-only and non-IR-aware baselines, with consistent gains across multiple backbones and supervision settings.

Significance. If the claims are appropriately scoped to synthetic infrared-style imagery, FusionRS is a large and potentially useful resource for studying modality alignment and modality-specific textual supervision in remote sensing vision-language models. The within-distribution experiments are coherent: the multi-backbone improvement in Table 14 is consistent, and the C-mixed retrieval pattern is a useful empirical finding. The authors also provide detailed construction statistics, qualitative examples, and a candid Limitations section. However, the paper currently overstates the contribution by implying general RGB–IR vision-language learning, while the two load-bearing pillars—synthetic-to-real transfer and the validity of IR-aware caption evaluation—are neither tested nor independently verified. The dataset is best presented as a large-scale synthetic infrared-style benchmark, not as a demonstrated foundation for real sensor-captured infrared VLM learning.

major comments (3)
  1. [Abstract & §4.4, Limitations] The abstract states that evaluation covers 'sensor-captured transfer,' but no experiment in the body uses any real sensor-captured RGB–IR remote sensing pair. VEDAI, DroneVehicle, M3FD, FLIR, and LLVIP appear only in Table 1 as related-work comparisons, and the evaluation protocol in §4.4 describes only FusionRS retrieval and generation. The final Limitations sentence concedes that effectiveness on real sensor-captured RGB–IR data 'still requires further validation.' This is an internal inconsistency that must be resolved: either add real-pair transfer experiments or remove the sensor-captured transfer claim and explicitly frame all conclusions as applying to synthetic infrared-style imagery.
  2. [§3.3, Table 11; §4.3, Table 5] The headline IR-aware captioning and IR-cue QA results are circular. The same Qwen2.5-VL-72B pipeline (RGB image + translated IR image + original caption, with the prompt template in Table 11) generated both the B-iraware training targets and the 10,000 IR-aware test captions used as evaluation references. The high Caption IR and IR-cue QA scores in Table 5 therefore measure how well the fine-tuned model mimics the annotation generator's style, not whether the model grounds infrared cues in physically meaningful ways. An independent evaluation—at minimum human judgments on a sample, or testing on real infrared captions—is needed to support the claim that IR-aware captions are the key to infrared-language grounding.
  3. [§3.2 and overall central claim] The load-bearing premise is that DiffV2IR's synthetic translations are a serviceable proxy for real infrared remote sensing observations. This is never tested: all training, retrieval, and captioning experiments are on the same synthetic distribution produced by the same translation model. If the synthetic-to-real gap is large, the claimed 'RGB–IR alignment' reduces to alignment with a grayscale stylization. The paper itself acknowledges this in Limitations. A small real-pair transfer evaluation (e.g., retrieval or captioning on VEDAI, DroneVehicle, or a similar RGB–IR remote sensing benchmark) would directly address this concern. Without it, the title and abstract should be revised to state clearly that the dataset and claims concern infrared-style synthetic images.
minor comments (5)
  1. [Title and throughout] The manuscript alternates between 'RGB–Infrared' and 'infrared-style.' Given the synthetic construction, the title and key claims should consistently use 'infrared-style' unless real-pair transfer is demonstrated.
  2. [§4.4 vs Table 4] The evaluation protocol says retrieval covers RGB-to-text and text-to-RGB, but Table 4 reports only IR→Text, Text→IR, RGB→IR, and IR→RGB. Either add the missing retrieval results or adjust the protocol description.
  3. [Table 5 and Tables 15-16] The metrics 'Cap. Auto,' 'Cap. IR,' 'Cap. R-L,' and 'IR-cue QA' are not defined. The appendix gives task definitions but not the automatic metrics or scoring protocols. Please specify (e.g., CIDEr, BLEU, ROUGE-L, and how the IR-cue QA answers are scored).
  4. [§3.4 / Table 7] The caption-quality filtering rules are described qualitatively. For reproducibility, the exact thresholds for 'too short,' 'too long,' 'generic,' and 'weak infrared cues' should be reported, especially since only 416 IR-aware validation captions survive the filter.
  5. [Tables 12 and 14] It is unclear whether the A-original/B-iraware/C-mixed rows in Table 14 are fine-tuned from the 580K-only model or trained from scratch. State the starting checkpoint and the number of fine-tuning steps for each setting.

Circularity Check

1 steps flagged · score 6.0 of 10

IR-aware caption evaluation is circular: test references are generated by the same Qwen2.5-VL-72B pipeline that produced the training targets, so the reported captioning/VQA gains partly measure student–teacher self-agreement rather than independent IR understanding.

  1. fitted input called prediction [§3.3 (IR-aware caption generation), Table 5 (VLM ablation), Table 8 (annotation setup)]
    "we generate 49,068 IR-aware training captions and 10,000 IR-aware test captions. ... Qwen2.5-VL-72B-Instruct is used only for generating IR-aware captions, not as the fine-tuned VLM backbone."

    The 10,000 IR-aware test captions used as ground truth for the captioning and IR-cue QA metrics in Table 5 are generated by the same Qwen2.5-VL-72B model and the same prompt as the 49,068 training captions. A model fine-tuned on teacher outputs will naturally score higher on teacher-generated references (e.g., Caption Auto 90.04, IR-cue QA 89.70) than a model trained on original captions, because the evaluation labels come from the same distribution the student was trained to imitate. This is not an independent ground truth; the improvement of B-iraware over A-original is partly a measure of student–teacher style agreement, not of genuine infrared-language competence. The paper's claim that IR-aware captions are crucial is therefore not validated against any external or sensor-captured ref

full rationale

The dataset construction itself (RGB-to-IR translation with DiffV2IR) is not circular: it is a stated design choice, and the limitations section transparently concedes that the synthetic IR is not validated on real sensor-captured RGB-IR pairs. The CLIP retrieval results (Table 4) use original captions from source datasets as external text supervision, so those numbers are not circular. The principal circularity is in the IR-aware caption track: the same Qwen2.5-VL-72B pipeline generates both the training targets (49,068 captions) and the test references (10,000 captions) with the same prompt, making the captioning/VQA benchmark a teacher-student self-consistency test. This inflates the reported gains of IR-aware supervision and weakens the claim that IR-aware captions improve task-conditioned infrared description. The paper's own Limitations statement further notes that effectiveness on real sensor-captured RGB-infrared paired data still requires validation, which is an external-validity gap rather than a circular step. Overall, the central dataset contribution is not circular, but one of the three headline claims (dual-modal captioning improvement) is substantially undermined by the self-generated evaluation labels, warranting a score of 6.

Assumptions & free parameters 3 free parameters · 3 assumptions · 1 invented entities

The dataset's value proposition rests on two external generators: DiffV2IR produces the entire IR modality from RGB inputs, and Qwen2.5-VL-72B produces the IR-aware captions that serve as both supervision and evaluation reference. Neither is independently validated in the paper (no real-IR transfer test, no human-verified caption test set). Because the central claims are empirical rather than derivational, the paper's free parameters are its hand-set filtering thresholds, prompt template, and loss-weight choices rather than fitted coefficients; the model-training hyperparameters are standard. Split integrity across overlapping sources (RSICD/RSITMD) is an additional untested assumption.

free parameters (3)
  • IR-aware caption filtering thresholds = not reported
    §3.4: cutoffs for "too short," "overly long," "weak infrared cues," "unsupported physical claims" are hand-set but never quantified; they determine which 49,068 training and 416 validation captions survive, and hence all IR-aware results.
  • IR-aware caption prompt template = see Table 11
    §3.3/Table 11: the annotation prompt instructs the generator to emphasize exactly the cues (grayscale intensity, high contrast, bright structures, dark/low-texture regions, structural outlines) that the later IR-cue QA test asks about, coupling label generation to the evaluation questions.
  • Tri-modal contrastive loss weights = 1/3 each for L_r,c, L_v,c, L_r,v (Eq. 4)
    §4.1 Eq. (4): equal weighting of RGB–text, IR–text, and RGB–IR losses is asserted without ablation; all retrieval results depend on this hand-set choice.
assumptions (3)
  • domain assumption DiffV2IR's translated "infrared-style" image is a faithful enough proxy for real infrared observations of remote sensing scenes that alignment learned on it transfers to physical IR.
    Invoked in §3.2 ("we convert RGB remote sensing images into infrared-style observations using DiffV2IR ... preserving the original scene layout and object-level semantics"). The entire IR modality is generated, and the paper's own Limitations admit real-sensor effectiveness "still requires further validation." If this fails, the dataset reduces to RGB→grayscale-stylization alignment.
  • domain assumption Qwen2.5-VL-72B-generated captions constitute valid ground truth for evaluating IR captioning and IR-cue QA, with no human-verified test reference required.
    §3.3/§3.4: IR-aware captions are filtered only by automatic quality checks plus sampled manual inspection; Table 5 and the abstract treat agreement with these captions (Caption Auto=90.04, IR-cue QA=89.70 under B-iraware) as evidence of improved infrared description.
  • domain assumption Sample-level splitting across five sources keeps test splits strictly held out, i.e., no near-duplicate imagery crosses train/test.
    §3.1: "The split is performed at the sample level," and the paper calls captioning/VQA "strictly held-out" (abstract). But RSITMD is built from RSICD imagery (their Table 1), and RGB translator inputs are not deduplicated across sources, so identical source photos could land in both train and test.
invented entities (1)
  • Synthetic "infrared-style" image modality
    purpose: The entire IR branch of the dataset, all IR-related training, and all IR-related evaluation claims rest on these generated images as a stand-in for real infrared observations.
    Produced by DiffV2IR (§3.2) from RGB inputs; no experiment in the paper connects them to physical IR statistics, and the paper's own Limitations disclaim sensor equivalency. The abstract's claim that evaluation covers "sensor-captured transfer" is contradicted by the body.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FusionRS: A Large-Scale RGB-Infrared-Style Remote Sensing Dataset for Cross-Modal Vision-Language Learning." pith.science (2026). https://pith.science/paper/IIAFWVPS

@misc{pith2026260617020,
  author       = {Pith},
  title        = {Pith review of: FusionRS: A Large-Scale RGB-Infrared-Style Remote Sensing Dataset for Cross-Modal Vision-Language Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IIAFWVPS}},
  note         = {Machine review of arXiv:2606.17020}
}
read the original abstract

Remote sensing vision-language models have advanced Earth observation, but available large-scale vision-language resources remain RGB-centered, leaving complementary infrared information underexplored. Infrared observations provide distinctive intensity structures, object boundaries, and illumination-invariant cues that complement conventional RGB imagery, yet large-scale RGB-infrared-text resources remain scarce. We introduce FusionRS, the first large-scale RGB-infrared-style-text dataset for controlled dual-modal remote sensing vision-language learning. It contains 600,000 spatially aligned pairs created by translating diverse public RGB remote sensing images into infrared-style counterparts. Each pair retains a conventional scene caption, and a curated subset adds 45,913 IR-aware captions describing observable intensity, contrast, texture, and structure while preserving scene semantics. We train CLIP-style models for RGB-infrared-style-text alignment and adapt a generative vision-language model with mixed task-conditioned caption supervision. Evaluation covers cross-modal retrieval, scaling and supervision ablations, sensor-captured transfer, and strictly held-out captioning and VQA. FusionRS substantially improves RGB-infrared-style alignment and infrared-to-text retrieval over RGB-only and non-IR-aware settings. Ablations show that IR-aware captions improve task-conditioned infrared description, demonstrating the value of modality-specific supervision. FusionRS provides a scalable foundation for controlled RGB-infrared remote sensing vision-language learning.

Figures

Figures reproduced from arXiv: 2606.17020 by the authors.

Figure 1
Figure 1. Overview of the proposed dual-modal RGB-IR-text remote sensing dataset construction pipeline. FusionRS [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Visual examples from the FusionRS dataset. For each scene, we show the original RGB image (left), [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Overview of the FusionRS construction and training pipeline. RGB remote sensing images are translated [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: CLIP IR-aware fine-tuning ablation across [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 3 linked inside Pith

  1. [1]

    Shuai Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, Sibo Song, Kai Dang, Peng Wang, Shijie Wang, Jun Tang, Humen Zhong, Yuanzhi Zhu, Mingkun Yang, Zhaohai Li, Jianqiang Wan, Pengfei Wang, Wei Ding, Zheren Fu, Yiheng Xu, and 8 others. 2025. https://arxiv.org/abs/2502.13923 Qwen2.5-vl technical report . Preprint, arXiv:2502.13923

    arXiv 2025

  2. [2]

    Yakoub Bazi, Laila Bashmal, Mohamad Mahmoud Al Rahhal, Riccardo Ricci, and Farid Melgani. 2024. Rs-llava: A large vision-language model for joint captioning and question answering in remote sensing imagery. Remote Sensing, 16(9):1477

    2024

  3. [3]

    Jinghao Cao, Xiajun Liu, and Rui Xue. 2026. Firemm-ir: An infrared-enhanced multi-modal large language model for comprehensive scene understanding in remote sensing forest fire monitoring. Sensors, 26(2):390

    2026

  4. [4]

    Zhe Cao, Jin Zhang, and Ruiheng Zhang. 2025. Irgpt: Understanding real-world infrared image with bi-cross-modal curriculum on large-scale benchmark. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 166--176

    2025

  5. [5]

    Gong Cheng, Junwei Han, and Xiaoqiang Lu. 2017. Remote sensing image scene classification: Benchmark and state of the art. Proceedings of the IEEE, 105(10):1865--1883

    2017

  6. [6]

    Mehdi Cherti, Romain Beaumont, Ross Wightman, Mitchell Wortsman, Gabriel Ilharco, Cade Gordon, Christoph Schuhmann, Ludwig Schmidt, and Jenia Jitsev. 2023. Reproducible scaling laws for contrastive language-image learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2818--2829

    2023

  7. [7]

    Wenliang Dai, Junnan Li, Dongxu Li, Anthony Tiong, Junqi Zhao, Weisheng Wang, Boyang Li, Pascale N Fung, and Steven Hoi. 2023. Instructblip: Towards general-purpose vision-language models with instruction tuning. Advances in neural information processing systems, 36:49250--49267

    2023

  8. [8]

    Qishen Ha, Kohei Watanabe, Takumi Karasawa, Yoshitaka Ushiku, and Tatsuya Harada. 2017. Mfnet: Towards real-time semantic segmentation for autonomous vehicles with multi-spectral scenes. In 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 5108--5115. IEEE

    2017

Show all 32 references
  1. [9]

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Liang Wang, Weizhu Chen, and 1 others. 2022. Lora: Low-rank adaptation of large language models. Iclr, 1(2):3

    2022

  2. [10]

    Xinyu Jia, Chuang Zhu, Minzhen Li, Wenqi Tang, and Wenli Zhou. 2021. Llvip: A visible-infrared paired dataset for low-light vision. In Proceedings of the IEEE/CVF international conference on computer vision, pages 3496--3504

    2021

  3. [11]

    Shixin Jiang, Zerui Chen, Jiafeng Liang, Yanyan Zhao, Ming Liu, and Bing Qin. 2024. Infrared-llava: Enhancing understanding of infrared images in multi-modal large language models. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 8573--8591

    2024

  4. [12]

    Kartik Kuckreja, Muhammad Sohail Danish, Muzammal Naseer, Abhijit Das, Salman Khan, and Fahad Shahbaz Khan. 2024. Geochat: Grounded large vision-language model for remote sensing. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 27831--27840

    2024

  5. [13]

    Junnan Li, Dongxu Li, Silvio Savarese, and Steven Hoi. 2023. Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models. In International conference on machine learning, pages 19730--19742. PMLR

    2023

  6. [14]

    Fan Liu, Delong Chen, Zhangqingyun Guan, Xiaocong Zhou, Jiale Zhu, Qiaolin Ye, Liyong Fu, and Jun Zhou. 2024 a . Remoteclip: A vision language foundation model for remote sensing. IEEE Transactions on Geoscience and Remote Sensing, 62:1--16

    2024

  7. [15]

    Haotian Liu, Chunyuan Li, Yuheng Li, and Yong Jae Lee. 2024 b . Improved baselines with visual instruction tuning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 26296--26306

    2024

  8. [16]

    Haotian Liu, Chunyuan Li, Yuheng Li, Bo Li, Yuanhan Zhang, Sheng Shen, and Yong Jae Lee. 2024 c . Llavanext: Improved reasoning, ocr, and world knowledge

    2024

  9. [17]

    Jinyuan Liu, Xin Fan, Zhanbo Huang, Guanyao Wu, Risheng Liu, Wei Zhong, and Zhongxuan Luo. 2022. Target-aware dual adversarial learning and a multi-scenario multi-modality benchmark to fuse infrared and visible for object detection. In Proceedings of the IEEE/CVF conference on...

    2022

  10. [18]

    Sylvain Lobry, Diego Marcos, Jesse Murray, and Devis Tuia. 2020. Rsvqa: Visual question answering for remote sensing data. IEEE Transactions on Geoscience and Remote Sensing, 58(12):8555--8566

    2020

  11. [19]

    Xiaoqiang Lu, Binqiang Wang, Xiangtao Zheng, and Xuelong Li. 2017. Exploring models and data for remote sensing image caption generation. IEEE Transactions on Geoscience and Remote Sensing, 56(4):2183--2195

    2017

  12. [20]

    Mehdi Moshtaghi, Siavash H Khajavi, and Joni Pajarinen. 2025. Rgb-th-bench: A dense benchmark for visual-thermal understanding of vision language models. arXiv preprint arXiv:2503.19654

    2025 arXiv

  13. [21]

    Chao Pang, Xingxing Weng, Jiang Wu, Jiayu Li, Yi Liu, Jiaxing Sun, Weijia Li, Shuai Wang, Litong Feng, Gui-Song Xia, and 1 others. 2025. Vhm: Versatile and honest vision language model for remote sensing image analysis. In Proceedings of the AAAI Conference on Artificial Intel...

    2025

  14. [22]

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, and 1 others. 2021. Learning transferable visual models from natural language supervision. In International conference on machine...

    2021

  15. [23]

    Lingyan Ran, Lidong Wang, Guangcong Wang, Peng Wang, and Yanning Zhang. 2025. Diffv2ir: visible-to-infrared diffusion model via vision-language understanding. arXiv preprint arXiv:2503.19012

    2025 arXiv

  16. [24]

    Sebastien Razakarivony and Frederic Jurie. 2016. Vehicle detection in aerial imagery: A small target detection benchmark. Journal of Visual Communication and Image Representation, 34:187--203

    2016

  17. [25]

    Yiming Sun, Bing Cao, Pengfei Zhu, and Qinghua Hu. 2022. Drone-based rgb-infrared cross-modality vehicle detection via uncertainty-aware learning. IEEE Transactions on Circuits and Systems for Video Technology, 32(10):6700--6713

    2022

  18. [26]

    Teledyne FLIR . 2018. Teledyne FLIR ADAS Dataset: FLIR Thermal Dataset for Algorithm Training . https://oem.flir.com/en-gb/solutions/automotive/adas-dataset-form/. Accessed: 2026-05-25

    2018

  19. [27]

    Peijin Wang, Huiyang Hu, Boyuan Tong, Ziqi Zhang, Fanglong Yao, Yingchao Feng, Zining Zhu, Hao Chang, Wenhui Diao, Qixiang Ye, and 1 others. 2024 a . Ringmogpt: A unified remote sensing foundation model for vision, language, and grounded tasks. IEEE Transactions on Geoscience ...

    2024

  20. [28]

    Zhecheng Wang, Rajanie Prabha, Tianyuan Huang, Jiajun Wu, and Ram Rajagopal. 2024 b . Skyscript: A large and semantically diverse vision-language dataset for remote sensing. In Proceedings of the AAAI Conference on Artificial Intelligence, 6, pages 5805--5813

    2024

  21. [29]

    Zhiqiang Yuan, Wenkai Zhang, Kun Fu, Xuan Li, Chubo Deng, Hongqi Wang, and Xian Sun. 2021. Exploring a fine-grained multiscale method for cross-modal remote sensing image retrieval. IEEE Transactions on Geoscience and Remote Sensing, 60:1--19

    2021

  22. [30]

    Zilun Zhang, Tiancheng Zhao, Yulong Guo, and Jianwei Yin. 2024. Rs5m and georsclip: A large-scale vision-language dataset and a large vision-language model for remote sensing. IEEE Transactions on Geoscience and Remote Sensing, 62:1--23

    2024

  23. [31]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list...

  24. [32]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.