Pith. sign in

REVIEW 3 major objections 7 minor 1 cited by

RSFAKE-1M: A Large-Scale Dataset for Detecting Diffusion-Generated Remote Sensing Forgeries

T0 review · 3 major / 7 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read RSFAKE-1M is a one-million-image benchmark for detecting diffusion-generated remote sensing forgeries; the paper shows that current detectors struggle on it, while training on it improves cross-dataset generalization.

desk verdict A genuinely useful large-scale diffusion forgery benchmark for remote sensing, but the fake real gap is confounded with source dataset (CLIP-RS vs fMoW), which muddies both headline claims. read the letter →

arxiv 2505.23283 v1 pith:PM3GT243 submitted 2025-05-29 cs.CV

Zhihong Tan , Jiayi Wang , Huiying Shi , Binyuan Huang , Hongchen Wei , Zhenzhong Chen This is my paper
classification cs.CV
keywords remotesensingforgerydetectiondiffusionmodelssyntheticimagebenchmarkdatasetcross-domaingeneralizationsatelliteimagerydeepfakegeography
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 sets out to show that forged satellite images made by modern diffusion models are a real and unsolved detection problem, and that a large, diverse training set is what detectors need to become robust. To that end it builds RSFAKE-1M, a million-image benchmark pairing 500,000 diffusion-generated fakes with 500,000 real remote sensing images, and runs a controlled comparison of existing detectors against uniform baselines. If the dataset holds up, it gives the community a shared test bed that is harder than earlier GAN-focused benchmarks and better at teaching cross-domain generalization. The stakes are practical: forged satellite imagery can mislead urban planning, environmental monitoring, and security decisions.

What carries the argument

The machinery is the dataset itself, RSFAKE-1M: 500,000 fake images produced by ten publicly available diffusion models (DiffusionSat-256/512, GeoRSSD, SDFRS, GeoSynth-text/sam/canny, CRSDiff, MapSat, RSPaint) fine-tuned on remote sensing data and covering six generation conditions—text prompts, segmentation masks, HED and Canny edge maps, vector maps, and inpainting masks—paired with 500,000 real images cropped from fMoW under a resolution-matching protocol that mirrors the fake set's size distribution. To make results comparable, the paper fixes a uniform standard ResNet-50 convolutional backbone with a two-layer MLP head and identical training hyperparameters across all datasets and subsets, so observed differences can be attributed to training data rather than architecture or tuning.

What would settle it

Train a baseline on RSFAKE-1M and test it on forgeries from a diffusion model not among the ten, generated with prompts derived from fMoW captions; if accuracy falls to near chance, the generalization claim is tied to the covered generators rather than to diffusion forgeries in general. A second check: if the same baseline can separate fMoW real crops from unmodified CLIP-RS images at high accuracy, then the benchmark's fake-real gap is partly dataset-of-origin bias.

Watch

Extended reading notes

Core claim

The central claim is that diffusion-generated remote sensing forgeries are a distinct and still-open detection problem, and that a large, diverse training corpus is the key to making detectors robust. On RSFAKE-1M's test split, the best pretrained general-purpose detector (UniversalFakeDetect) reaches about 61% accuracy and an area under the ROC curve of 0.67, while remote-sensing-specific detectors such as Geo-DefakeHop and two FLDCF variants fall to roughly 52-56% accuracy, close to chance. A uniform standard ResNet-50 convolutional backbone trained on RSFAKE-1M reaches 99.88% in-domain accuracy and 72.97% average accuracy across five datasets, outperforming the same baseline trained on FSI, Geo-DefakeHop, DM-AER, or FLDCF. Within the dataset, text-only training transfers best across generation conditions, and scaling the training set from 10% to 100% monotonically improves generalization and robustness to downsampling, JPEG compression, and blur.

Load-bearing premise

The benchmark's value depends on the ten diffusion models and six generation conditions being representative of the forgeries a real detector would face, and on the fake-real differences coming from actual forgery cues rather than from the different source datasets behind the fake and real images.

Editorial extensions

If this is right

  • Existing remote sensing and general-purpose AIGC detectors cannot be deployed directly on diffusion-generated satellite forgeries; the best pretrained detector reaches only about 61% accuracy and 0.67 AUC on RSFAKE-1M.
  • Training on RSFAKE-1M transfers across domains: the unified baseline reaches 72.97% average accuracy over five datasets, beating models trained on FSI, Geo-DefakeHop, DM-AER, or FLDCF.
  • Text-only forged images are the most transferable training signal across generation conditions, so future datasets should keep caption-driven generation at the core.
  • Dataset scale is a direct lever: increasing training data from 10% to 100% of RSFAKE-1M improves accuracy on out-of-domain data and under downsampling, JPEG compression, and Gaussian blur.
  • The dataset's ten-model, six-condition splits expose that generation condition changes can matter as much as architecture changes, since GeoSynth variants transfer unevenly to each other.

Reading between the lines

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

  • A stronger stress test the paper does not run: hold out an entire modern diffusion family, retrain on RSFAKE-1M, and check whether accuracy survives; the cross-generator tables suggest it may drop sharply.
  • If a detector trained on RSFAKE-1M can also separate fMoW crops from raw CLIP-RS images, part of the measured difficulty is source-dataset bias rather than genuine forgery signatures.
  • The text-only advantage hints that adding more caption-driven generators, and adding conditions like layout or editing tools, would stretch the benchmark's difficulty in the directions real forgers use.
  • The scaling curves imply RSFAKE-1M could serve as a pretraining corpus for a universal satellite-forgery detector, with fine-tuning on smaller application-specific sets as the final step.
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 / 7 minor

Summary. The paper introduces RSFAKE-1M, a large-scale dataset for detecting diffusion-generated remote sensing forgeries, containing 500K fake images produced by ten diffusion models under six generation conditions (text, segmentation masks, edge maps, vector maps, and inpainting) and 500K real images cropped from the fMoW dataset. The authors evaluate existing forgery detectors on RSFAKE-1M, train unified ResNet-50 baselines for cross-dataset generalization, analyze intra-dataset generalization across generators and conditions, and study the effect of training-set scale on robustness. They conclude that RSFAKE-1M is challenging for current detectors and improves generalization and robustness when used for training.

Significance. If valid, RSFAKE-1M is a substantial resource: it is significantly larger and more diverse than existing remote sensing forgery benchmarks, focuses on diffusion generators, and is publicly released. The multi-generator, multi-condition design is a strength, and the paper provides a systematic comparison that can inform future detector development. However, the central empirical claims rest on the assumption that the fake/real label reflects manipulation rather than source-dataset differences, and the cross-dataset advantage claimed for RSFAKE-1M is less clear when self-test performance is excluded. The dataset itself could not be inspected at review time, so reproducibility could not be independently verified.

major comments (3)
  1. [Section 3.1 vs. 3.2; Tables 3 and 4] The fake images are generated from the top 450K CLIP-RS image-text pairs (Section 3.1), while the real images are cropped from fMoW (Section 3.2). These sources differ systematically in sensor type, geographic coverage, and curation (including the CLIP-Score filter), so the fake/real label may be confounded with source dataset rather than manipulation. The low accuracy of pretrained detectors on RSFAKE-1M in Table 3, and the cross-dataset gains in Table 4, could therefore reflect detection of 'CLIP-RS-like versus fMoW-like' content instead of genuine forgery cues. No control experiment isolates this factor. I recommend adding an experiment that includes real CLIP-RS images (or fakes generated from fMoW-derived prompts) to assess the magnitude of the source-domain confound, or at minimum a classifier trained to distinguish CLIP-RS from fMoW to quantify how separable the two source domains are.
  2. [Table 4] The claim that models trained on RSFAKE-1M 'generally outperform' models trained on other datasets in cross-dataset evaluation is not supported when the self-test is excluded. The reported average of 72.97 for RSFAKE-1M is driven by its 99.88 on the RSFAKE-1M test set; the leave-one-out average over the other four datasets is 66.24, lower than the FLDCF-trained model's leave-one-out average of 72.81. RSFAKE-1M-trained models are also weaker on FSI, Geo-DefakeHop, and FLDCF test sets. Please report per-dataset results with the self-test excluded and apply a statistical test over multiple training runs.
  3. [Tables 3-6 and Figure 3] All experiments report single-run accuracies without error bars or significance tests. Many cross-generator and cross-condition differences in Tables 5 and 6 are a few percentage points, and claims such as 'text-only forgeries provide stronger generalization signals' (Section 4.3) or 'models trained on larger subsets consistently achieve higher accuracy' (Section 4.4) would be more credible with multiple seeds and confidence intervals or paired significance tests.
minor comments (7)
  1. [Section 1] The sentence 'The development of forgery detectors typically depends on the availability of suitable datasets' is repeated verbatim twice in the introduction; remove the duplicate.
  2. [Throughout] The name of the DiffusionSat model is inconsistently capitalized ('DiffusionSat' vs. 'DiffusionSAT'); please standardize.
  3. [Table 2] Table 2 states that 'All models are trained on remote sensing data,' but MapSat is trained on OpenStreetMap data from Central Belt Scotland; clarify the distinction between full remote sensing training corpora and region- or task-specific training sets.
  4. [Figure 1] The figure caption contains a typo: 'Diffuison Model' should be 'Diffusion Model'.
  5. [Table 1] The FID values reported in Table 1 are not accompanied by details of how they were computed (e.g., which feature extractor and which reference set were used); please specify this in the text or a footnote.
  6. [Section 3.2] The phrase 'not explicitly paired' in the real-image collection description only rules out exact image-level pairing; it does not rule out distributional non-independence. Please discuss the distributional relationship between CLIP-RS and fMoW more explicitly.
  7. [Data availability] The dataset URL could not be verified at review time. Please ensure the dataset, generation code, and a data card (including license and intended uses) are publicly accessible and stable at the time of publication.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: RSFAKE-1M is an empirical benchmark paper whose claims are supported by external evaluations, not by a derivation that reduces to its own inputs.

full rationale

The paper makes no analytic derivation whose conclusion is equivalent to its assumptions. Its central claims are empirical: that existing detectors score poorly on RSFAKE-1M, that models trained on RSFAKE-1M generalize better cross-dataset, and that larger training subsets improve robustness. These claims are supported by held-out test splits, external benchmark datasets (FSI, Geo-DefakeHop, DM-AER, FLDCF), and fixed ResNet-50 baselines with published hyperparameters. There is no fitted parameter that is later renamed as a prediction, and no uniqueness theorem or prior self-citation is invoked to force a modeling choice. The only overlap with author identity is the use of CLIP-RS [38] as the prompt/source-image pool for fake generation; that is a data-sourcing choice, not a load-bearing mathematical premise, and the real-image set is drawn independently from fMoW [42]. The skeptic's concern that the fake/real labels may correlate with source dataset (CLIP-RS-derived generations versus fMoW crops) is a validity and generalizability risk, not a circularity: recognizing that confound would be an empirical failure mode of the benchmark, not a logical reduction of its conclusions to its inputs. No circular step can be exhibited from the paper's own equations or construction protocol, so the appropriate finding is no significant circularity.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

The ledger lists the manual design choices that shape the benchmark. None of them are fitted to a target result, but they influence the difficulty and generalization properties that the paper reports. The axioms are domain assumptions about the quality and representativeness of the source data.

free parameters (5)
  • class balance = 50% real / 50% fake
    The paper intentionally uses a 1:1 ratio to avoid class imbalance (Section 3.2). This choice affects training dynamics and the difficulty of the benchmark.
  • generation condition mix = 50% text, 40% image-guided, 10% inpainting
    Manual allocation across the six conditions (Figure 2c) determines the diversity of the fake set and influences the intra-dataset generalization results.
  • resolution mix = 70% 512, 20% 256, 10% 768
    The real-image sampling protocol (Section 3.2) matches the resolution distribution of the fake images, a deliberate choice to avoid resolution artifacts.
  • CLIP-Score filtering threshold = top 450k of 10M pairs
    The paper retains the top 450,000 image-text pairs from CLIP-RS based on CLIP-Score (Section 3.1). This threshold is a manual choice that affects prompt quality and diversity.
  • inpainting mask size = 25% random square masks
    Used for the RSPaint inpainting condition (Section 3.1). A different mask size would change the detection difficulty.
assumptions (3)
  • domain assumption The ten selected diffusion models were indeed trained or fine-tuned on remote sensing data and are publicly available.
    The paper selects models 'that have been trained or fine-tuned on remote sensing imagery' (Section 3.1). The validity of this assumption is taken from the cited model papers and is not independently verified here.
  • domain assumption The CLIP-RS captions, refined with GeoChat, are semantically correct descriptions of the source images.
    The text prompts for generation are derived from these captions (Section 3.1). If captions are noisy, the generated scenes may not align with the intended distribution.
  • domain assumption fMoW is an appropriate proxy for real remote sensing imagery in terms of geographic and resolution diversity.
    The real images are cropped from fMoW (Section 3.2). If fMoW is not representative of the scenes in the fake set, detectors may learn dataset-of-origin cues rather than forgery cues.

how reviews work

0 comments
Cite this review

Pith. "Pith review of RSFAKE-1M: A Large-Scale Dataset for Detecting Diffusion-Generated Remote Sensing Forgeries." pith.science (2026). https://pith.science/paper/PM3GT243

@misc{pith2026250523283,
  author       = {Pith},
  title        = {Pith review of: RSFAKE-1M: A Large-Scale Dataset for Detecting Diffusion-Generated Remote Sensing Forgeries},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PM3GT243}},
  note         = {Machine review of arXiv:2505.23283}
}
read the original abstract

Detecting forged remote sensing images is becoming increasingly critical, as such imagery plays a vital role in environmental monitoring, urban planning, and national security. While diffusion models have emerged as the dominant paradigm for image generation, their impact on remote sensing forgery detection remains underexplored. Existing benchmarks primarily target GAN-based forgeries or focus on natural images, limiting progress in this critical domain. To address this gap, we introduce RSFAKE-1M, a large-scale dataset of 500K forged and 500K real remote sensing images. The fake images are generated by ten diffusion models fine-tuned on remote sensing data, covering six generation conditions such as text prompts, structural guidance, and inpainting. This paper presents the construction of RSFAKE-1M along with a comprehensive experimental evaluation using both existing detectors and unified baselines. The results reveal that diffusion-based remote sensing forgeries remain challenging for current methods, and that models trained on RSFAKE-1M exhibit notably improved generalization and robustness. Our findings underscore the importance of RSFAKE-1M as a foundation for developing and evaluating next-generation forgery detection approaches in the remote sensing domain. The dataset and other supplementary materials are available at https://huggingface.co/datasets/TZHSW/RSFAKE/.

Figures

Figures reproduced from arXiv: 2505.23283 by the authors.

Figure 1
Figure 1. Overview of the RSFAKE-1M dataset generation pipeline. The dataset consists of 500,000 fake images generated by 10 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Dataset composition of the collected synthetic remote sensing image corpus. From left to right: (a) train/ [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Generalization and robustness evaluation of baseline models with ResNet-50 backbone trained on varying portions of [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Towards a satellite image manipulation and deepfake localization benchmark dataset

    cs.CV 2026-08 conditional novelty 5.0 of 10

    A 60-image satellite benchmark with ground-truth masks and metadata for testing manipulation detection and localization in remote sensing imagery.

Reference graph

Works this paper leans on

46 extracted references · 42 canonical work pages · cited by 1 Pith paper

  1. [1]

    Deep fake geography? when geospatial data encounter artificial intelligence.Cartography and Geographic Information Science, 48(4):338–352, 2021

    Bo Zhao, Shaozeng Zhang, Chunxue Xu, Yifan Sun, and Chengbin Deng. Deep fake geography? when geospatial data encounter artificial intelligence.Cartography and Geographic Information Science, 48(4):338–352, 2021

    work page 2021

  2. [2]

    Hong-Shuo Chen, Kaitai Zhang, Shuowen Hu, Suya You, and C. C. Jay Kuo. Geo-defakehop: High-performance geographic fake image detection. 2021

    work page 2021

  3. [3]

    DM-AER-DeepFake-V1 dataset, 2022

    DeepMedia. DM-AER-DeepFake-V1 dataset, 2022. URL https://github.com/RijulGupta-DM/ deepfake-satellite-images

    work page 2022

  4. [4]

    Fldcf: A collaborative framework for forgery localization and detection in satellite imagery.IEEE Transactions on Geo- science and Remote Sensing, 2024

    Jialu Sui, Ding Ma, C-C Jay Kuo, and Man-On Pun. Fldcf: A collaborative framework for forgery localization and detection in satellite imagery.IEEE Transactions on Geo- science and Remote Sensing, 2024

    work page 2024

  5. [5]

    Denoising diffusion probabilistic models.Advances in Neural Infor- mation Processing Systems, 33:6840–6851, 2020

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models.Advances in Neural Infor- mation Processing Systems, 33:6840–6851, 2020

    work page 2020

  6. [6]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. High-resolution image synthesis with latent diffusion models. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10684–10695, 2022

    2022

  7. [7]

    Av-deepfake1m: A large-scale llm-driven audio-visual deepfake dataset

    Zhixi Cai, Shreya Ghosh, Aman Pankaj Adatia, Munawar Hayat, Abhinav Dhall, Tom Gedeon, and Kalin Stefanov. Av-deepfake1m: A large-scale llm-driven audio-visual deepfake dataset. InProceedings of the 32nd ACM In- ternational Conference on Multimedia, pages 7414–7423, New York, NY , USA, 2024. Association for Computing Machinery. ISBN 9798400706868

    work page 2024

  8. [8]

    Dire for diffusion-generated image detection

    Zhendong Wang, Jianmin Bao, Wengang Zhou, Weilun Wang, Hezhen Hu, Hong Chen, and Houqiang Li. Dire for diffusion-generated image detection. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 22445–22455, 2023

    2023

Show all 46 references
  1. [9]

    Advanc- ing generalized deepfake detector with forgery perception guidance

    Ruiyang Xia, Dawei Zhou, Decheng Liu, Lin Yuan, Shuodi Wang, Jie Li, Nannan Wang, and Xinbo Gao. Advanc- ing generalized deepfake detector with forgery perception guidance. InProceedings of the 32nd ACM International Conference on Multimedia, pages 6676–6685. Association for C...

    2024

  2. [10]

    Urban green space planning based on re- mote sensing and geographic information systems.Remote Sensing, 14(17):4213, 2022

    Hua Bai, Ziwei Li, Hanlong Guo, Haopeng Chen, and Pingping Luo. Urban green space planning based on re- mote sensing and geographic information systems.Remote Sensing, 14(17):4213, 2022

    2022

  3. [11]

    Remote sensing big data for water envi-6 ronment monitoring: Current status, challenges, and future prospects.Earth’s Future, 10(2):e2021EF002289, 2022

    Jinyue Chen, Shuisen Chen, Rao Fu, Dan Li, Hao Jiang, Chongyang Wang, Yongshi Peng, Kai Jia, and Bren- dan J Hicks. Remote sensing big data for water envi-6 ronment monitoring: Current status, challenges, and future prospects.Earth’s Future, 10(2):e2021EF002289, 2022

    2022

  4. [12]

    Remote sensing of irrigated agriculture: Oppor- tunities and challenges.Remote sensing, 2(9):2274–2304, 2010

    Mutlu Ozdogan, Yang Yang, George Allez, and Chelsea Cervantes. Remote sensing of irrigated agriculture: Oppor- tunities and challenges.Remote sensing, 2(9):2274–2304, 2010

    2010

  5. [13]

    Hypersectral imaging for military and security applica- tions: Combining myriad processing and sensing tech- niques.IEEE Geoscience and Remote Sensing Magazine, 7(2):101–117, 2019

    Michal Shimoni, Rob Haelterman, and Christiaan Perneel. Hypersectral imaging for military and security applica- tions: Combining myriad processing and sensing tech- niques.IEEE Geoscience and Remote Sensing Magazine, 7(2):101–117, 2019

    2019

  6. [14]

    A style-based generator architecture for generative adversarial networks

    Tero Karras, Samuli Laine, and Timo Aila. A style-based generator architecture for generative adversarial networks. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4401–4410, 2019

    2019

  7. [15]

    Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A. Efros. Unpaired image-to-image translation using cycle- consistent adversarial networks. In2017 IEEE Interna- tional Conference on Computer Vision, pages 2242–2251, 2017

    2017

  8. [16]

    Analyzing and improv- ing the image quality of stylegan

    Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Analyzing and improv- ing the image quality of stylegan. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8110–8119, 2020

    2020

  9. [17]

    Diffusion models beat GANs on image synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion models beat GANs on image synthesis. volume 34, pages 8780– 8794, 2021

    2021

  10. [18]

    Wildfake: A large-scale and hierarchical dataset for ai-generated images detection

    Yan Hong, Jianming Feng, Haoxing Chen, Jun Lan, Huijia Zhu, Weiqiang Wang, and Jianfu Zhang. Wildfake: A large-scale and hierarchical dataset for ai-generated images detection. InProceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 3500–3508, 2025

    2025

  11. [19]

    Genimage: A million-scale bench- mark for detecting ai-generated image.Advances in Neural Information Processing Systems, 36:77771–77782, 2023

    Mingjian Zhu, Hanting Chen, Qiangyu Yan, Xudong Huang, Guanyu Lin, Wei Li, Zhijun Tu, Hailin Hu, Jie Hu, and Yunhe Wang. Genimage: A million-scale bench- mark for detecting ai-generated image.Advances in Neural Information Processing Systems, 36:77771–77782, 2023

    2023

  12. [20]

    Artifact: A large-scale dataset with artificial and factual images for generalizable and robust synthetic im- age detection

    Md Awsafur Rahman, Bishmoy Paul, Najibul Haque Sarker, Zaber Ibn Abdul Hakim, and Shaikh Anowarul Fattah. Artifact: A large-scale dataset with artificial and factual images for generalizable and robust synthetic im- age detection. InIEEE International Conference on Image Proce...

    2023

  13. [21]

    Wang, Evan Montoya, David Munechika, Haoyang Yang, Benjamin Hoover, and Duen Horng Chau

    Zijie J. Wang, Evan Montoya, David Munechika, Haoyang Yang, Benjamin Hoover, and Duen Horng Chau. Dif- fusiondb: A large-scale prompt gallery dataset for text- to-image generative models. InAnnual Meeting of the Association for Computational Linguistics, 2022

    2022

  14. [22]

    Holistically-nested edge detection

    Saining Xie and Zhuowen Tu. Holistically-nested edge detection. InProceedings of the IEEE international con- ference on computer vision, pages 1395–1403, 2015

    2015

  15. [23]

    A computational approach to edge detec- tion.IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-8(6):679–698, 1986

    John Canny. A computational approach to edge detec- tion.IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-8(6):679–698, 1986

    1986

  16. [24]

    Remote sensing image dataset expansion based on generative ad- versarial networks with modified shuffle attention.Sensors, 21(14), 2021

    Lu Chen, Hongjun Wang, and Xianghao Meng. Remote sensing image dataset expansion based on generative ad- versarial networks with modified shuffle attention.Sensors, 21(14), 2021. ISSN 1424-8220

    2021

  17. [25]

    Yadong Li, Sébastien Mavromatis, Feng Zhang, Zhenhong Du, Jean Sequeira, Zhongyi Wang, Xianwei Zhao, and Renyi Liu. Single-image super-resolution for remote sens- ing images using a deep generative adversarial network with local and global attention mechanisms.IEEE Transac- ti...

    2022

  18. [26]

    Meng Xu, Furong Deng, Sen Jia, Xiuping Jia, and An- tonio J. Plaza. Attention mechanism-based generative adversarial networks for cloud removal in landsat images. Remote Sensing of Environment, 271:112902, 2022. ISSN 0034-4257

    2022

  19. [27]

    Text-to-remote-sensing-image generation with structured generative adversarial networks

    Rui Zhao and Zhenwei Shi. Text-to-remote-sensing-image generation with structured generative adversarial networks. IEEE Geoscience and Remote Sensing Letters, 19:1–5, 2022

    2022

  20. [28]

    Remote sensing image synthesis via graphical generative adversarial networks

    Guangxing Wang, Guoshuai Dong, Hui Li, Lirong Han, Xuanwen Tao, and Peng Ren. Remote sensing image synthesis via graphical generative adversarial networks. In IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium, pages 10027–10030, 2019

    2019

  21. [29]

    Disastergan: Generative adversarial networks for remote sensing disaster image generation.Remote Sensing, 13 (21), 2021

    Xue Rui, Yang Cao, Xin Yuan, Yu Kang, and Weiguo Song. Disastergan: Generative adversarial networks for remote sensing disaster image generation.Remote Sensing, 13 (21), 2021. ISSN 2072-4292

    2021

  22. [30]

    Remote sensing image synthesis via semantic embed- ding generative adversarial networks.IEEE Transactions on Geoscience and Remote Sensing, 61:1–11, 2023

    Chendan Wang, Bowen Chen, Zhengxia Zou, and Zhenwei Shi. Remote sensing image synthesis via semantic embed- ding generative adversarial networks.IEEE Transactions on Geoscience and Remote Sensing, 61:1–11, 2023

    2023

  23. [31]

    Crs-diff: Controllable remote sensing image generation with diffusion model

    Datao Tang, Xiangyong Cao, Xingsong Hou, Zhongyuan Jiang, Junmin Liu, and Deyu Meng. Crs-diff: Controllable remote sensing image generation with diffusion model. IEEE Transactions on Geoscience and Remote Sensing, 62:1–14, 2024

    2024

  24. [32]

    Geosynth: Contextually-aware high- resolution satellite image synthesis

    Srikumar Sastry, Subash Khanal, Aayush Dhakal, and Nathan Jacobs. Geosynth: Contextually-aware high- resolution satellite image synthesis. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, pages 460–470, 2024

    2024

  25. [33]

    Diffusionsat: A generative foundation model for satellite imagery

    Samar Khanna, Patrick Liu, Linqi Zhou, Chenlin Meng, Robin Rombach, Marshall Burke, David Lobell, and Ste- fano Ermon. Diffusionsat: A generative foundation model for satellite imagery. 2024

    2024

  26. [34]

    Miguel Espinosa and Elliot J. Crowley. Generate your own scotland: Satellite image generation conditioned on maps. NeurIPS Workshop on Diffusion Models, 2023

    2023

  27. [35]

    Tackling few-shot segmentation in re- mote sensing via inpainting diffusion model

    Steve Andreas Immanuel, Woojin Cho, Junhyuk Heo, and Darongsae Kwon. Tackling few-shot segmentation in re- mote sensing via inpainting diffusion model. 2025

    2025

  28. [36]

    Efficient and controllable remote sensing fake sample generation based on diffusion model.IEEE Transactions on Geoscience and Remote Sensing, 61:1–12, 2023.7

    Zhiqiang Yuan, Chongyang Hao, Ruixue Zhou, Jialiang Chen, Miao Yu, Wenkai Zhang, Hongqi Wang, and Xian Sun. Efficient and controllable remote sensing fake sample generation based on diffusion model.IEEE Transactions on Geoscience and Remote Sensing, 61:1–12, 2023.7

    2023

  29. [37]

    Rs5m and georsclip: A large scale vision-language dataset and a large vision-language model for remote sens- ing.IEEE Transactions on Geoscience and Remote Sens- ing, 62:1–23, 2024

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

    2024

  30. [38]

    Remote sensing semantic segmentation quality assessment based on vision language model

    Huiying Shi, Zhihong Tan, Zhihan Zhang, Hongchen Wei, Yaosi Hu, Yingxue Zhang, and Zhenzhong Chen. Remote sensing semantic segmentation quality assessment based on vision language model. 2025

    2025

  31. [39]

    Geochat: Grounded large vision-language model for remote sensing

    Kartik Kuckreja, Muhammad Sohail Danish, Muzammal Naseer, Abhijit Das, Salman Khan, and Fahad Shahbaz Khan. Geochat: Grounded large vision-language model for remote sensing. InProceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, pages 27831–27840, 2024

    2024

  32. [40]

    Fast segment anything

    Xu Zhao, Wenchao Ding, Yongqi An, Yinglong Du, Tao Yu, Min Li, Ming Tang, and Jinqiao Wang. Fast segment anything. 2023

    2023

  33. [41]

    Planet dump

    OpenStreetMap contributors. Planet dump. https:// planet.osm.org, 2017

    2017

  34. [42]

    Functional map of the world

    Gordon Christie, Neil Fendley, James Wilson, and Ryan Mukherjee. Functional map of the world. InProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 6172–6180, 2018

    2018

  35. [43]

    Towards universal fake image detectors that generalize across gener- ative models

    Utkarsh Ojha, Yuheng Li, and Yong Jae Lee. Towards universal fake image detectors that generalize across gener- ative models. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023

    2023

  36. [44]

    A sanity check for ai- generated image detection

    Shilin Yan, Ouxiang Li, Jiayin Cai, Yanbin Hao, Xiaolong Jiang, Yao Hu, and Weidi Xie. A sanity check for ai- generated image detection. 2025

    2025

  37. [45]

    Improving synthetic image detection towards generalization: An image transformation perspec- tive

    Ouxiang Li, Jiayin Cai, Yanbin Hao, Xiaolong Jiang, Yao Hu, and Fuli Feng. Improving synthetic image detection towards generalization: An image transformation perspec- tive. 2025

    2025

  38. [46]

    Chuangchuang Tan, Renshuai Tao, Huan Liu, Guanghua Gu, Baoyuan Wu, Yao Zhao, and Yunchao Wei. C2p- clip: Injecting category common prompt in clip to enhance generalization in deepfake detection.Proceedings of the AAAI Conference on Artificial Intelligence, 39(7):7184– 7192, 2025. 8

    2025

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.