pith. sign in

arxiv: 2606.25312 · v1 · pith:YIG7QN3Xnew · submitted 2026-06-24 · 💻 cs.CV

LEVIRDet: A Million-Scale 159-Category Dataset and Foundation Model for Universal Remote Sensing Object Detection

Qinzhe Yang , Dongyu Wang , Haohan Niu , Jia Xu , Zhenwei Shi , Zhengxia Zou This is my paper

Pith reviewed 2026-06-25 21:28 UTC · model grok-4.3

classification 💻 cs.CV
keywords remote sensingobject detectionlarge-scale datasetfoundation modelzero-shot detectionmulti-level taxonomyground sampling distancecross-domain generalization
0
0 comments X

The pith

A new 159-category remote sensing dataset trains a model that detects objects zero-shot across nine external benchmarks, outperforming fully supervised methods by 5.02 mAP on average.

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

The paper presents LEVIRDet-159 as the largest remote sensing object detection dataset, spanning 159 categories, 2.56 million bounding boxes, and a multi-level taxonomy. From this data it constructs LEVIRDetNet, which predicts ground sampling distance on the fly and uses that prediction to modulate queries and allocate detections across granularity levels. The resulting model is tested under strict zero-shot conditions on nine separate benchmarks drawn from different sensors and resolutions. It records state-of-the-art scores on each, exceeding the strongest fully supervised competing method by an average of 5.02 mAP under the benchmark's own primary metric.

Core claim

LEVIRDet-159 supplies 159 categories, 2.56 million bounding boxes, and 700k fine-grained annotations under a multi-level taxonomy, exceeding prior datasets by factors of roughly 7x in images, 6x in instances, and 4x in categories. LEVIRDetNet adds online visual GSD prediction, GSD-conditioned query modulation and allocation, and a hierarchy-aware detection head. Without any target-domain training or fine-tuning, the model reaches state-of-the-art detection performance on nine external benchmarks and improves the strongest fully supervised competitors by 5.02 mAP on average under each benchmark's primary metric.

What carries the argument

LEVIRDetNet, which couples online visual Ground Sampling Distance prediction, GSD-conditioned query modulation and allocation, and a hierarchy-aware detection head for mixed-granularity supervision.

If this is right

  • A single trained detector can be deployed across multiple sensors and resolutions without per-domain retraining.
  • Multi-level taxonomy supervision enables detection at both coarse and fine category granularities within one forward pass.
  • Online GSD prediction removes the need for manual scale metadata at inference time.
  • Cross-domain generalization reduces the cost of building new remote sensing detection systems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same GSD-conditioning pattern could be tested on non-remote-sensing imagery where object scale varies strongly with acquisition distance.
  • Releasing the dataset and weights allows direct measurement of whether further scaling the number of categories produces additional zero-shot gains.
  • The hierarchy-aware head suggests a route for unifying detection and fine-grained classification in one architecture.

Load-bearing premise

The nine external benchmarks used for zero-shot evaluation are sufficiently diverse in sensors, resolutions, and category systems to demonstrate universal generalization across real-world remote sensing conditions.

What would settle it

Performance of the released model measured on a tenth remote sensing benchmark whose sensor type, resolution range, or category system lies outside the training distribution and the nine evaluation sets; if the average gain falls below the reported 5.02 mAP or drops below the supervised baseline, the universal claim is falsified.

Figures

Figures reproduced from arXiv: 2606.25312 by Dongyu Wang, Haohan Niu, Jia Xu, Qinzhe Yang, Zhengxia Zou, Zhenwei Shi.

Figure 1
Figure 1. Figure 1: Overview of target-training-free cross-benchmark performance and dataset-scale comparison. (a) Dataset-scale comparison for [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Comparison of dataset statistics. The three panels compare image counts by object number, object counts by object size, and [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: An image source distribution of LEVIRDet-159, covering [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Category and annotation expansion of LEVIRDet-159. The figure compares category coverage before and after unification, [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: An annotation refinement composition of LEVIRDet-159. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: A category-wise visualization of LEVIRDet-159, showing representative object crops across all categories. [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: An overview of the proposed LEVIRDetNet. The framework adapts a DEIM-style end-to-end detector to universal remote [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Visualization of LEVIRDetNet on unseen external remote sensing images outside LEVIRDet-159. Each image is processed by [PITH_FULL_IMAGE:figures/full_fig_p013_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: A comparison between LEVIRDetNet and open-set/ [PITH_FULL_IMAGE:figures/full_fig_p014_9.png] view at source ↗
read the original abstract

Remote sensing object detection has advanced rapidly with the development of large-scale benchmarks and modern detection architectures. However, existing datasets and detectors remain fragmented. Most benchmarks focus on limited categories, fixed spatial resolutions, or a single sensor, while detectors still struggle to work across different sensors and categorical systems. In this paper, we introduce LEVIRDet-159, the largest and most comprehensive remote sensing object detection dataset to date, with 159 categories, 2.56 million bounding boxes, and 700k fine-grained annotations under a multi-level taxonomy. In each key scale dimension, LEVIRDet-159 exceeds the corresponding largest existing remote sensing object detection dataset, containing approximately (7x) more images, (6x) more object instances, and (4x) more categories. Based on this dataset, we design LEVIRDetNet, a scale-hierarchy-aware detection foundation model for universal remote sensing object detection. LEVIRDetNet couples online visual Ground Sampling Distance (GSD) prediction, GSD-conditioned query modulation and allocation, and a hierarchy-aware detection head for mixed-granularity remote sensing supervision. Under stringent evaluation settings, LEVIRDetNet demonstrates strong cross-domain generalization. Even without target-domain training or fine-tuning, it achieves state-of-the-art detection performance on 9 external benchmarks, improving the strongest fully supervised competing methods by 5.02 mAP on average under each benchmark's primary metric. We hope this study will facilitate the development of strongly generalizable remote sensing object detection across diverse category systems, spatial resolutions, and sensor platforms. The dataset and trained models will be released at https://qinzheyang.github.io/LEVIRDet/, accompanying the final paper.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper introduces LEVIRDet-159, the largest remote sensing object detection dataset to date (159 categories, 2.56 million bounding boxes, 700k fine-grained annotations under a multi-level taxonomy), exceeding prior datasets by factors of ~7x images, ~6x instances, and ~4x categories. It also presents LEVIRDetNet, a scale-hierarchy-aware foundation model incorporating online GSD prediction, GSD-conditioned query modulation/allocation, and a hierarchy-aware detection head. The central claim is that, without any target-domain training or fine-tuning, LEVIRDetNet achieves SOTA zero-shot detection on 9 external benchmarks and improves the strongest fully supervised competing methods by 5.02 mAP on average under each benchmark's primary metric.

Significance. If the zero-shot generalization claim holds after verification of benchmark diversity and training details, the work would be significant for the field: it supplies both a large-scale, multi-taxonomy dataset and a model architecture explicitly designed for cross-sensor, cross-resolution, and cross-category remote sensing detection, potentially reducing reliance on per-domain fine-tuning. The planned public release of the dataset and models is a clear strength.

major comments (2)
  1. [Abstract] Abstract: The claim that LEVIRDetNet demonstrates 'strong cross-domain generalization' and 'universal remote sensing object detection' across 'diverse category systems, spatial resolutions, and sensor platforms' is load-bearing for the 5.02 mAP average improvement result. However, the manuscript supplies no quantitative breakdown (sensor distribution, GSD statistics, or category intersection sizes) of the 9 external benchmarks relative to LEVIRDet-159. Without this, it is impossible to confirm that the benchmarks collectively span meaningfully different conditions rather than predominantly similar high-resolution optical sources.
  2. [Methods/Evaluation (implied by absence)] The provided manuscript text contains no training protocols, data-overlap checks between LEVIRDet-159 and the 9 external benchmarks, ablation studies on the GSD-prediction or hierarchy-aware components, or error analysis. These elements are required to evaluate whether the reported zero-shot gains are robust or could be explained by data leakage or benchmark similarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for highlighting the need for explicit evidence of benchmark diversity and evaluation rigor to support the zero-shot claims. We will revise the manuscript to incorporate the requested quantitative breakdowns, protocols, and analyses, which will strengthen the presentation without altering the core results.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The claim that LEVIRDetNet demonstrates 'strong cross-domain generalization' and 'universal remote sensing object detection' across 'diverse category systems, spatial resolutions, and sensor platforms' is load-bearing for the 5.02 mAP average improvement result. However, the manuscript supplies no quantitative breakdown (sensor distribution, GSD statistics, or category intersection sizes) of the 9 external benchmarks relative to LEVIRDet-159. Without this, it is impossible to confirm that the benchmarks collectively span meaningfully different conditions rather than predominantly similar high-resolution optical sources.

    Authors: We agree that the current abstract and main text lack a consolidated quantitative breakdown of benchmark diversity. In the revision we will add a dedicated table (new Table 2) reporting, for each of the 9 external benchmarks: primary sensor type(s), GSD range statistics, number of images/instances, and category intersection size with LEVIRDet-159 (both exact and hierarchical matches). This table will also include a summary row showing the aggregate coverage across optical, SAR, and multi-spectral sources to demonstrate that the benchmarks are not dominated by high-resolution optical imagery similar to the training distribution. revision: yes

  2. Referee: [Methods/Evaluation (implied by absence)] The provided manuscript text contains no training protocols, data-overlap checks between LEVIRDet-159 and the 9 external benchmarks, ablation studies on the GSD-prediction or hierarchy-aware components, or error analysis. These elements are required to evaluate whether the reported zero-shot gains are robust or could be explained by data leakage or benchmark similarity.

    Authors: The full manuscript contains a Methods section describing the overall training setup, but we acknowledge that explicit data-overlap verification, component ablations, and error analysis are insufficiently detailed. We will expand Section 4 (Experiments) with: (i) a data-overlap analysis confirming zero instance-level leakage via image-hash and geographic-coordinate checks; (ii) ablation tables isolating the contribution of online GSD prediction and the hierarchy-aware head; and (iii) a qualitative error analysis highlighting failure modes across sensor/GSD regimes. These additions will be placed before the main zero-shot results to allow readers to assess robustness directly. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical zero-shot results on distinct external benchmarks

full rationale

The paper trains LEVIRDetNet on the newly introduced LEVIRDet-159 dataset and reports mAP improvements on nine explicitly external benchmarks under zero-shot conditions with no target-domain training or fine-tuning. This performance claim is an empirical measurement on held-out data and does not reduce by construction to any fitted parameter, self-definition, or self-citation chain within the paper. No equations, uniqueness theorems, or ansatzes are shown that would make the reported 5.02 mAP average gain tautological with the training inputs. The central derivation chain (dataset construction → model design → cross-benchmark evaluation) remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract alone supplies no explicit free parameters, mathematical axioms, or new postulated entities; model components are described at high level only.

pith-pipeline@v0.9.1-grok · 5867 in / 1172 out tokens · 26373 ms · 2026-06-25T21:28:33.514062+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

94 extracted references · 9 linked inside Pith

  1. [1]

    Object detection in aerial images: A large-scale benchmark and challenges,

    J. Ding, N. Xue, G.-S. Xia, X. Bai, W. Yang, M. Y . Yang, S. Belongie, J. Luo, M. Datcu, M. Pelillo, and L. Zhang, “Object detection in aerial images: A large-scale benchmark and challenges,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 11, pp. 7778– 7796, 2022

    2022

  2. [2]

    Object detection in 20 years: A survey,

    Z. Zou, K. Chen, Z. Shi, Y . Guo, and J. Ye, “Object detection in 20 years: A survey,”Proceedings of the IEEE, vol. 111, no. 3, pp. 257–276, 2023

    2023

  3. [3]

    Dota: A large-scale dataset for object detection in aerial images,

    G.-S. Xia, X. Bai, J. Ding, Z. Zhu, S. Belongie, J. Luo, M. Datcu, M. Pelillo, and L. Zhang, “Dota: A large-scale dataset for object detection in aerial images,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 3974–3983

    2018

  4. [4]

    Object detection in optical remote sensing images: A survey and a new benchmark,

    K. Li, G. Wan, G. Cheng, L. Meng, and J. Han, “Object detection in optical remote sensing images: A survey and a new benchmark,”ISPRS journal of photogrammetry and remote sensing, vol. 159, pp. 296–307, 2020

    2020

  5. [5]

    Fair1m: A benchmark dataset for fine-grained object recognition in high-resolution remote sensing imagery,

    X. Sun, P. Wang, Z. Yan, F. Xu, R. Wang, W. Diao, J. Chen, J. Li, Y . Feng, T. Xuet al., “Fair1m: A benchmark dataset for fine-grained object recognition in high-resolution remote sensing imagery,”ISPRS Journal of Photogrammetry and Remote Sensing, vol. 184, pp. 116–130, 2022

    2022

  6. [6]

    Towards large-scale small object detection: Survey and benchmarks,

    G. Cheng, X. Yuan, X. Yao, K. Yan, Q. Zeng, X. Xie, and J. Han, “Towards large-scale small object detection: Survey and benchmarks,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 11, pp. 13 467–13 488, 2023

    2023

  7. [7]

    A degraded recon- struction enhancement-based method for tiny ship detection in remote sensing images with a new large-scale dataset,

    J. Chen, K. Chen, H. Chen, Z. Zou, and Z. Shi, “A degraded recon- struction enhancement-based method for tiny ship detection in remote sensing images with a new large-scale dataset,”IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1–14, 2022

    2022

  8. [8]

    Orientation robust object detection in aerial images using deep convolutional neural network,

    H. Zhu, X. Chen, W. Dai, K. Fu, Q. Ye, and J. Jiao, “Orientation robust object detection in aerial images using deep convolutional neural network,” in2015 IEEE International Conference on Image Processing (ICIP), 2015, pp. 3735–3739

    2015

  9. [9]

    Multi-class geospatial object detection and geographic image classification based on collection of part detectors,

    G. Cheng, J. Han, P. Zhou, and L. Guo, “Multi-class geospatial object detection and geographic image classification based on collection of part detectors,”ISPRS Journal of Photogrammetry and Remote Sensing, vol. 98, pp. 119–132, 2014

    2014

  10. [10]

    Accurate object localization in remote sensing images based on convolutional neural networks,

    Y . Long, Y . Gong, Z. Xiao, and Q. Liu, “Accurate object localization in remote sensing images based on convolutional neural networks,”IEEE Transactions on Geoscience and Remote Sensing, vol. 55, no. 5, pp. 2486–2498, 2017

    2017

  11. [11]

    Hierarchical and robust convolutional neural network for very high-resolution remote sensing object detection,

    Y . Zhang, Y . Yuan, Y . Feng, and X. Lu, “Hierarchical and robust convolutional neural network for very high-resolution remote sensing object detection,”IEEE Transactions on Geoscience and Remote Sensing, vol. 57, no. 8, pp. 5535–5548, 2019

    2019

  12. [12]

    Learning spatial context: Using stuffto find things,

    G. Heitz and D. Koller, “Learning spatial context: Using stuffto find things,” inEuropean conference on computer vision. Springer, 2008, pp. 30–43

    2008

  13. [13]

    A high resolution optical satellite image dataset for ship recognition and some new baselines,

    Z. Liu, L. Yuan, L. Weng, and Y . Yang, “A high resolution optical satellite image dataset for ship recognition and some new baselines,” inInternational conference on pattern recognition applications and methods, vol. 2. SciTePress, 2017, pp. 324–331

    2017

  14. [14]

    Vehicle detection in aerial imagery: A small target detection benchmark,

    S. Razakarivony and F. Jurie, “Vehicle detection in aerial imagery: A small target detection benchmark,”Journal of Visual Communication and Image Representation, vol. 34, pp. 187–203, 2016

    2016

  15. [15]

    Vehicle detection in aerial images,

    M. Y . Yang, W. Liao, X. Li, Y . Cao, and B. Rosenhahn, “Vehicle detection in aerial images,”Photogrammetric Engineering&Remote Sensing, vol. 85, no. 4, pp. 297–304, 2019

    2019

  16. [16]

    Fast multiclass vehicle detection on aerial images,

    K. Liu and G. Mattyus, “Fast multiclass vehicle detection on aerial images,”IEEE Geoscience and Remote Sensing Letters, vol. 12, no. 9, pp. 1938–1942, 2015

    1938

  17. [17]

    A tool for bridge detection in major infrastructure works using satellite images,

    K. Nogueira, C. Cesar, P. H. Gama, G. L. Machado, and J. A. dos Santos, “A tool for bridge detection in major infrastructure works using satellite images,” in2019 XV Workshop de Vis˜ ao Computacional (WVC). IEEE, 2019, pp. 72–77

    2019

  18. [18]

    Overhead imagery research data set—an annotated data library & tools to aid in the development of computer vision algorithms,

    F. Tanner, B. Colder, C. Pullen, D. Heagy, M. Eppolito, V . Carlan, C. Oertel, and P. Sallee, “Overhead imagery research data set—an annotated data library & tools to aid in the development of computer vision algorithms,” in2009 IEEE Applied Imagery Pattern Recognition Workshop (AIPR 2009). IEEE, 2009, pp. 1–8

    2009

  19. [19]

    Mar20: A benchmark for military aircraft recognition in remote sensing images,

    Y . Wenqi, C. Gong, W. Meijun, Y . Yanqing, X. Xingxing, Y . Xiwen, and H. Junwei, “Mar20: A benchmark for military aircraft recognition in remote sensing images,”National Remote Sensing Bulletin, vol. 27, no. 12, pp. 2688–2696, 2024

    2024

  20. [20]

    Vhrships: An extensive bench- mark dataset for scalable deep learning-based ship detection applica- tions,

    S. Kızılkaya, U. Alganci, and E. Sertel, “Vhrships: An extensive bench- mark dataset for scalable deep learning-based ship detection applica- tions,”ISPRS International Journal of Geo-Information, vol. 11, no. 8, p. 445, 2022

    2022

  21. [21]

    Shiprsimagenet: A large-scale fine-grained dataset for ship detection in high-resolution optical remote sensing images,

    Z. Zhang, L. Zhang, Y . Wang, P. Feng, and R. He, “Shiprsimagenet: A large-scale fine-grained dataset for ship detection in high-resolution optical remote sensing images,”IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 14, pp. 8458– 8472, 2021

    2021

  22. [22]

    A new spatial-oriented object detection framework for remote sensing images,

    D. Yu and S. Ji, “A new spatial-oriented object detection framework for remote sensing images,”IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1–16, 2021

    2021

  23. [23]

    A benchmark dataset for deep learning-based airplane detection: Hrplanes,

    T. Bakırman and E. Sertel, “A benchmark dataset for deep learning-based airplane detection: Hrplanes,”International Journal of Engineering and Geosciences, vol. 8, no. 3, pp. 212–223, 2023

    2023

  24. [24]

    Drone-based object counting by spatially regularized regional proposal network,

    M.-R. Hsieh, Y .-L. Lin, and W. H. Hsu, “Drone-based object counting by spatially regularized regional proposal network,” inProceedings of the IEEE international conference on computer vision, 2017, pp. 4145–4153

    2017

  25. [25]

    Vhrv: Very high- resolution benchmark dataset for vessel detection,

    F. B ¨uy¨ukkanber, M. Yanalak, and N. Musao ˘glu, “Vhrv: Very high- resolution benchmark dataset for vessel detection,”Remote Sensing Applications: Society and Environment, vol. 39, p. 101641, 2025

    2025

  26. [26]

    Performance evolution of yolo models in remote sensing images,

    I. Hassan and Z. Xinyou, “Performance evolution of yolo models in remote sensing images,” in2024 21st International Computer Confer- ence on Wavelet Active Media Technology and Information Processing (ICCWAMTIP). IEEE, 2024, pp. 1–4

    2024

  27. [27]

    Com- plex optical remote-sensing aircraft detection dataset and benchmark,

    T. Shi, J. Gong, S. Jiang, X. Zhi, G. Bao, Y . Sun, and W. Zhang, “Com- plex optical remote-sensing aircraft detection dataset and benchmark,” IEEE Transactions on Geoscience and Remote Sensing, vol. 61, pp. 1–9, 2023

    2023

  28. [28]

    A benchmark dataset for aircraft detection in optical remote sensing imagery,

    J. Hu, X. Zhi, B. Zhang, T. Shi, Q. Cui, and X. Sun, “A benchmark dataset for aircraft detection in optical remote sensing imagery,”Remote Sensing, vol. 16, no. 24, p. 4699, 2024

    2024

  29. [29]

    Efficient remote sensing instance segmentation with linear-time state space distilled visual foun- dation models,

    Q. Yang, K. Chen, J. Xu, Z. Shi, and Z. Zou, “Efficient remote sensing instance segmentation with linear-time state space distilled visual foun- dation models,”IEEE Transactions on Geoscience and Remote Sensing, pp. 1–1, 2026

    2026

  30. [30]

    Faster r-cnn: Towards real-time object detection with region proposal networks,

    S. Ren, K. He, R. Girshick, and J. Sun, “Faster r-cnn: Towards real-time object detection with region proposal networks,”IEEE transactions on pattern analysis and machine intelligence, vol. 39, no. 6, pp. 1137–1149, 2016

    2016

  31. [31]

    Cascade r-cnn: Delving into high quality object detection,

    Z. Cai and N. Vasconcelos, “Cascade r-cnn: Delving into high quality object detection,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 6154–6162

    2018

  32. [32]

    You only look once: Unified, real-time object detection,

    J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You only look once: Unified, real-time object detection,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 779– 788

    2016

  33. [33]

    Ssd: Single shot multibox detector,

    W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y . Fu, and A. C. Berg, “Ssd: Single shot multibox detector,” inEuropean conference on computer vision. Springer, 2016, pp. 21–37. JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. XX, XXX XXXX 17

    2016

  34. [34]

    Focal loss for dense object detection,

    T.-Y . Lin, P. Goyal, R. Girshick, K. He, and P. Doll ´ar, “Focal loss for dense object detection,” inProceedings of the IEEE international conference on computer vision, 2017, pp. 2980–2988

    2017

  35. [35]

    Fcos: Fully convolutional one- stage object detection,

    Z. Tian, C. Shen, H. Chen, and T. He, “Fcos: Fully convolutional one- stage object detection,” inProceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 9627–9636

    2019

  36. [36]

    Objects as points,

    X. Zhou, D. Wang, and P. Kr¨ahenb¨uhl, “Objects as points,”arXiv preprint arXiv:1904.07850, 2019

    Pith/arXiv arXiv 1904

  37. [37]

    End-to-end object detection with transformers,

    N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, “End-to-end object detection with transformers,” inEu- ropean conference on computer vision. Springer, 2020, pp. 213–229

    2020

  38. [38]

    Deformable detr: Deformable transformers for end-to-end object detection,

    X. Zhu, W. Su, L. Lu, B. Li, X. Wang, and J. Dai, “Deformable detr: Deformable transformers for end-to-end object detection,”arXiv preprint arXiv:2010.04159, 2020

    Pith/arXiv arXiv 2010

  39. [39]

    Dino: Detr with improved denoising anchor boxes for end-to-end object detection,

    H. Zhang, F. Li, S. Liu, L. Zhang, H. Su, J. Zhu, L. M. Ni, and H.-Y . Shum, “Dino: Detr with improved denoising anchor boxes for end-to-end object detection,”arXiv preprint arXiv:2203.03605, 2022

    Pith/arXiv arXiv 2022

  40. [40]

    Detrs beat yolos on real-time object detection,

    Y . Zhao, W. Lv, S. Xu, J. Wei, G. Wang, Q. Dang, Y . Liu, and J. Chen, “Detrs beat yolos on real-time object detection,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2024, pp. 16 965–16 974

    2024

  41. [41]

    Deim: Detr with improved matching for fast convergence,

    S. Huang, Z. Lu, X. Cun, Y . Yu, X. Zhou, and X. Shen, “Deim: Detr with improved matching for fast convergence,” inProceedings of the computer vision and pattern recognition conference, 2025, pp. 15 162–15 171

    2025

  42. [42]

    Grounding dino: Marrying dino with grounded pre-training for open-set object detection,

    S. Liu, Z. Zeng, T. Ren, F. Li, H. Zhang, J. Yang, Q. Jiang, C. Li, J. Yang, H. Suet al., “Grounding dino: Marrying dino with grounded pre-training for open-set object detection,” inEuropean conference on computer vision. Springer, 2024, pp. 38–55

    2024

  43. [43]

    Locate anything on earth: Advancing open-vocabulary object detection for remote sensing community,

    J. Pan, Y . Liu, Y . Fu, M. Ma, J. Li, D. P. Paudel, L. Van Gool, and X. Huang, “Locate anything on earth: Advancing open-vocabulary object detection for remote sensing community,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 39, no. 6, 2025, pp. 6281– 6289

    2025

  44. [44]

    Simple open-vocabulary object detection,

    M. Minderer, A. Gritsenko, A. Stone, M. Neumann, D. Weissenborn, A. Dosovitskiy, A. Mahendran, A. Arnab, M. Dehghani, Z. Shenet al., “Simple open-vocabulary object detection,” inEuropean conference on computer vision. Springer, 2022, pp. 728–755

    2022

  45. [45]

    Learning to holistically detect bridges from large-size vhr remote sensing imagery,

    Y . Li, J. Luo, Y . Zhang, Y . Tan, J.-G. Yu, and S. Bai, “Learning to holistically detect bridges from large-size vhr remote sensing imagery,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 46, no. 12, pp. 11 507–11 523, 2024

    2024

  46. [46]

    Rareplanes: Synthetic data takes flight,

    J. Shermeyer, T. Hossler, A. Van Etten, D. Hogan, R. Lewis, and D. Kim, “Rareplanes: Synthetic data takes flight,” inProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2021, pp. 207–217

    2021

  47. [47]

    A benchmark data set for aircraft type recognition from remote sensing images,

    Z. Wu, S. Wan, X. Wang, M. Tan, L. Zou, X. Li, and Y . Chen, “A benchmark data set for aircraft type recognition from remote sensing images,”Applied Soft Computing, vol. 89, p. 106132, 2020

    2020

  48. [48]

    Feature pyramid networks for object detection,

    T.-Y . Lin, P. Doll´ar, R. Girshick, K. He, B. Hariharan, and S. Belongie, “Feature pyramid networks for object detection,” in2017 IEEE Confer- ence on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 936–944

    2017

  49. [49]

    Detecting tiny objects in aerial images: A normalized wasserstein distance and a new benchmark,

    C. Xu, J. Wang, , W. Yang, H. Yu, L. Yu, and G.-S. Xia, “Detecting tiny objects in aerial images: A normalized wasserstein distance and a new benchmark,”ISPRS Journal of Photogrammetry and Remote Sensing, 2022

    2022

  50. [50]

    Scrdet++: Detecting small, cluttered and rotated objects via instance-level feature denoising and rotation loss smoothing,

    X. Yang, J. Yan, W. Liao, X. Yang, J. Tang, and T. He, “Scrdet++: Detecting small, cluttered and rotated objects via instance-level feature denoising and rotation loss smoothing,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 2, pp. 2384–2399, 2023

    2023

  51. [51]

    Target detection in hyperspectral remote sensing image: Current status and challenges,

    B. Chen, L. Liu, Z. Zou, and Z. Shi, “Target detection in hyperspectral remote sensing image: Current status and challenges,”Remote Sensing, vol. 15, no. 13, p. 3223, 2023

    2023

  52. [52]

    State space models meet remote sensing: A survey,

    Q. Yang, C. Liu, J. Xu, Z. Shi, and Z. Zou, “State space models meet remote sensing: A survey,”SCIENCE CHINA Information Sciences, 2026

    2026

  53. [53]

    xview: Objects in context in overhead imagery,

    D. Lam, R. Kuzma, K. McGee, S. Dooley, M. Laielli, M. Klaric, Y . Bulatov, and B. McCord, “xview: Objects in context in overhead imagery,”arXiv preprint arXiv:1802.07856, 2018

    Pith/arXiv arXiv 2018

  54. [54]

    Use of the stair vision library within the isprs 2d semantic labeling benchmark (vaihingen),

    I. Markus Gerke, “Use of the stair vision library within the isprs 2d semantic labeling benchmark (vaihingen),”Use of the stair vision library within the isprs 2d semantic labeling benchmark (vaihingen), 2014

    2014

  55. [55]

    Functional map of the world,

    G. Christie, N. Fendley, J. Wilson, and R. Mukherjee, “Functional map of the world,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018, pp. 6172–6180

    2018

  56. [56]

    Automatic ship classification from optical aerial images with convolutional neural networks,

    A.-J. Gallego, A. Pertusa, and P. Gil, “Automatic ship classification from optical aerial images with convolutional neural networks,”Remote Sensing, vol. 10, no. 4, p. 511, 2018

    2018

  57. [57]

    Airbus ship detection challenge,

    Airbus, “Airbus ship detection challenge,”Kaggle, 2018

    2018

  58. [58]

    Airbus aircraft detection challenge,

    ——, “Airbus aircraft detection challenge,”Kaggle, 2021

    2021

  59. [59]

    Gliding vertex on the horizontal bounding box for multi-oriented object detection,

    Y . Xu, M. Fu, Q. Wang, Y . Wang, K. Chen, G.-S. Xia, and X. Bai, “Gliding vertex on the horizontal bounding box for multi-oriented object detection,”IEEE Transactions on Pattern Analysis and Machine Intelli- gence, vol. 43, no. 4, pp. 1452–1459, 2021

    2021

  60. [60]

    Multi-oriented object detection in aerial images with double horizontal rectangles,

    G. Nie and H. Huang, “Multi-oriented object detection in aerial images with double horizontal rectangles,”IEEE Transactions on Pattern Analy- sis and Machine Intelligence, vol. 45, no. 4, pp. 4932–4944, 2023

    2023

  61. [61]

    Mmdetection: Open mmlab detection toolbox and benchmark,

    K. Chen, J. Wang, J. Pang, Y . Cao, Y . Xiong, X. Li, S. Sun, W. Feng, Z. Liu, J. Xuet al., “Mmdetection: Open mmlab detection toolbox and benchmark,”arXiv preprint arXiv:1906.07155, 2019

    Pith/arXiv arXiv 1906

  62. [62]

    MMEngine: Openmmlab foundational library for train- ing deep learning models,

    M. Contributors, “MMEngine: Openmmlab foundational library for train- ing deep learning models,” 2022

    2022

  63. [63]

    A large contextual dataset for classification, detection and counting of cars with deep learning,

    T. N. Mundhenk, G. Konjevod, W. A. Sakla, and K. Boakye, “A large contextual dataset for classification, detection and counting of cars with deep learning,” inEuropean conference on computer vision. Springer, 2016, pp. 785–800

    2016

  64. [64]

    Can semantic labeling methods generalize to any city? the inria aerial image labeling benchmark,

    E. Maggiori, Y . Tarabalka, G. Charpiat, and P. Alliez, “Can semantic labeling methods generalize to any city? the inria aerial image labeling benchmark,” in2017 IEEE International geoscience and remote sensing symposium (IGARSS). IEEE, 2017, pp. 3226–3229

    2017

  65. [65]

    Bigearthnet: A large-scale benchmark archive for remote sensing image understanding,

    G. Sumbul, M. Charfuelan, B. Demir, and V . Markl, “Bigearthnet: A large-scale benchmark archive for remote sensing image understanding,” arXiv preprint arXiv:1902.06148, 2019

    arXiv 1902

  66. [66]

    Deep high-resolution representation learning for visual recognition,

    J. Wang, K. Sun, T. Cheng, B. Jiang, C. Deng, Y . Zhao, D. Liu, Y . Mu, M. Tan, X. Wanget al., “Deep high-resolution representation learning for visual recognition,”IEEE transactions on pattern analysis and machine intelligence, vol. 43, no. 10, pp. 3349–3364, 2020

    2020

  67. [67]

    Fast r-cnn,

    R. Girshick, “Fast r-cnn,” inProceedings of the IEEE international conference on computer vision, 2015, pp. 1440–1448

    2015

  68. [68]

    You only look one-level feature,

    Q. Chen, Y . Wang, T. Yang, X. Zhang, J. Cheng, and J. Sun, “You only look one-level feature,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 13 039–13 048

    2021

  69. [69]

    Sparse r-cnn: An end-to-end framework for object detection,

    P. Sun, R. Zhang, Y . Jiang, T. Kong, C. Xu, W. Zhan, M. Tomizuka, Z. Yuan, and P. Luo, “Sparse r-cnn: An end-to-end framework for object detection,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 12, pp. 15 650–15 664, 2023

    2023

  70. [70]

    Rtmdet: An empirical study of designing real-time object detectors,

    C. Lyu, W. Zhang, H. Huang, Y . Zhou, Y . Wang, Y . Liu, S. Zhang, and K. Chen, “Rtmdet: An empirical study of designing real-time object detectors,”arXiv preprint arXiv:2212.07784, 2022

    arXiv 2022

  71. [71]

    Efficientdet: Scalable and efficient object detection,

    M. Tan, R. Pang, and Q. V . Le, “Efficientdet: Scalable and efficient object detection,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 10 781–10 790

    2020

  72. [72]

    Scale-aware trident networks for object detection,

    Y . Li, Y . Chen, N. Wang, and Z. Zhang, “Scale-aware trident networks for object detection,” inProceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 6054–6063

    2019

  73. [73]

    Yolox: Exceeding yolo series in 2021,

    Z. Ge, S. Liu, F. Wang, Z. Li, and J. Sun, “Yolox: Exceeding yolo series in 2021,”arXiv preprint arXiv:2107.08430, 2021

    Pith/arXiv arXiv 2021

  74. [74]

    Yolov12: Attention-centric real-time object detectors,

    Y . Tian, Q. Ye, and D. Doermann, “Yolov12: Attention-centric real-time object detectors,”Advances in neural information processing systems, vol. 38, pp. 78 433–78 457, 2026

    2026

  75. [75]

    You only look once based-c2fghost using efficient siou loss function for airplane detection,

    M.-T. Do, M.-H. Ha, D.-C. Nguyen, and O. T.-C. Chen, “You only look once based-c2fghost using efficient siou loss function for airplane detection,” in2024 9th International Conference on Frontiers of Signal Processing (ICFSP). IEEE, 2024, pp. 1–5

    2024

  76. [76]

    Real-time object detection meets dinov3,

    S. Huang, Y . Hou, L. Liu, X. Yu, and X. Shen, “Real-time object detection meets dinov3,”arXiv preprint arXiv:2509.20787, 2025

    arXiv 2025

  77. [77]

    Dynamicvis: An efficient and general visual foundation model for remote sensing image understanding,

    K. Chen, C. Liu, B. Chen, W. Li, Z. Zou, and Z. Shi, “Dynamicvis: An efficient and general visual foundation model for remote sensing image understanding,”arXiv preprint arXiv:2503.16426, 2025

    arXiv 2025

  78. [78]

    What is yolov5: A deep look into the internal features of the popular object detector,

    R. Khanam and M. Hussain, “What is yolov5: A deep look into the internal features of the popular object detector,”arXiv preprint arXiv:2407.20892, 2024

    arXiv 2024

  79. [79]

    Small object detection in unmanned aerial vehicle images using multi-scale hybrid attention,

    G. Song, H. Du, X. Zhang, F. Bao, and Y . Zhang, “Small object detection in unmanned aerial vehicle images using multi-scale hybrid attention,” Engineering Applications of Artificial Intelligence, vol. 128, p. 107455, 2024

    2024

  80. [80]

    Sim ´eoni, H

    O. Sim ´eoni, H. V . V o, M. Seitzer, F. Baldassarre, M. Oquab, C. Jose, V . Khalidov, M. Szafraniec, S. Yi, M. Ramamonjisoaet al., “Dinov3,” arXiv preprint arXiv:2508.10104, 2025

    Pith/arXiv arXiv 2025

Showing first 80 references.