Pith. sign in

REVIEW 5 major objections 6 minor 54 references

MMO-IG: Multi-Class and Multi-Scale Object Image Generation for Remote Sensing

T0 review · 5 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read MMO-IG claims that conditioning a diffusion model on a grayscale instance map plus a text count prompt yields remote sensing images whose object classes, locations, and scales are correct, and detectors trained on those images transfer to…

desk verdict A plausible and well-engineered pipeline for generating instance-labeled RS imagery, but the downstream gains are too thin and under-evaluated to justify the 'excellent performance' claim. read the letter →

arxiv 2412.13684 v3 pith:H43M6XDO submitted 2024-12-18 cs.CV

classification cs.CV
keywords remotesensingimagegenerationobjectdetectiondiffusionmodelinstance-levellabelsdataaugmentationsyntheticlayout-to-image
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

MMO-IG argues that remote sensing images with dense, instance-level detection labels can be synthesized rather than hand-annotated. The paper's claim is that a diffusion model conditioned on a grayscale iso-spacing instance map (ISIM) decodes each coded region into an object of the intended class, location, and scale, while a knowledge graph (SCDKG) and a textual object-count prompt (SODI) keep the layout and content realistic. If true, this gives object detectors an essentially unlimited supply of labeled training images, reducing the annotation bottleneck that limits remote sensing object detection. The paper supports the claim with generation-quality metrics and with downstream experiments where detectors trained on DIOR plus 20,000 generated images outperform their real-data-only baselines.

What carries the argument

The load-bearing object is the Iso-spacing Instance Map (ISIM): a grayscale control image in which each object class is assigned the gray value floor(255 m / M), each object instance is a filled region with that value, region centroids encode location, and region area and aspect ratio encode scale and shape. It is what makes instance-level labels a byproduct of generation. Two auxiliary mechanisms carry the rest: SCDKG, a directed class-interdependency graph plus per-class probability densities over coordinates, scales, and aspect ratios that synthesizes the ISIM layout, and SODI, a text prompt built by counting objects per class, which conditions the denoiser so generated content stays aligned with the ISIM. The decoder is a latent diffusion network initialized from the architecture of reference [10] and augmented with trainable residual blocks.

What would settle it

Train a standard detector such as YOLO on 20,000 MMO-IG-generated images alone, with no real DIOR images, and evaluate on the real DIOR test set; if its accuracy is far below a model trained on the real DIOR train split, the synthetic supervision is not carrying the claimed information. A finer check is per-class ISIM-to-image agreement to see whether the classes with detector losses are exactly the classes whose gray levels are confused with adjacent codes.

Watch

Extended reading notes

Core claim

The central discovery, as the authors state it, is that a single grayscale control map can carry all the information needed to generate a labeled remote sensing image: each object class is encoded as a distinct gray level, each object's location is the region centroid, and its scale and aspect ratio are the region's geometry. MMO-IG trains a latent diffusion decoder to render each ISIM region into the corresponding object, while a spatial-cross dependency knowledge graph built from DIOR statistics samples plausible class sequences, counts, locations, scales, and co-occurrences, and a structured object distribution instruction such as 'a remote sensing image with 7 airplanes and 4 vehicles' anchors the global scene. On DIOR and DIOR-R, the method reports class accuracy of 97.9% and 98.2%, count accuracy of 98.7% and 97.2%, FID of 34.48 and 35.07 versus 41.31 and 48.43 for the GLIGEN baseline, and per-class downstream gains up to 8.1 percentage points for five detectors.

Load-bearing premise

The load-bearing premise is that the diffusion decoder can resolve the small gray-level spacing between class codes (about 12-13 gray units for 20 classes), so every region in the ISIM is rendered as the intended object rather than a neighboring class.

Editorial extensions

If this is right

  • Every MMO-IG image is generated from an ISIM that contains the ground-truth class, location, and scale of every object, so synthetic labels come for free and can be used directly to train detectors.
  • Because SCDKG learns co-occurrence and geometry statistics from DIOR, generated scenes inherit realistic inter-class arrangements rather than the rigid layouts of prompt-based generators.
  • SODI lifts class and count agreement to about 98% and lowers FID and CAS relative to BLIP captioning, which means generated content and its instance-level labels stay aligned.
  • Augmenting the DIOR training set with 20,000 generated images improves R-CNN, Faster R-CNN, YOLO, PANet, and CornerNet on the real test set, with per-class gains up to 8.1 percentage points.
  • Because ISIM is a grayscale map, the same pipeline works for both horizontal (DIOR) and oriented (DIOR-R) annotation conventions.

Reading between the lines

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

  • A natural untested extension is to use ISIM as dense supervision for semantic or instance segmentation, since the control map is already a full per-pixel label image.
  • The 12-13 gray-level spacing between 20 classes means that moving to datasets with many more classes will press the class codes closer together; an adaptive multichannel encoding would be the direct next test.
  • The paper's own reported class and count mismatches of about 2% imply the synthetic labels contain a small amount of noise, so gains might increase if generated samples are filtered by agreement with a pre-trained detector.
  • Because the paper's limitation section notes that rare instances can be misrepresented, a targeted evaluation should check whether downstream gains are driven by common classes while rare classes such as windmill or toll station are left behind.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper proposes MMO-IG, a diffusion-based generator for remote sensing images with dense instance-level labels. It introduces three components: ISIM, a grayscale instance map that encodes each object's class, location, and scale; SCDKG, a probabilistic knowledge graph that models class co-occurrence and spatial geometric characteristics; and SODI, a structured text prompt that summarizes object counts and classes. The method is evaluated on DIOR and DIOR-R using FID, CAS, Accc, and Accn, and the synthetic images are used to augment training of five object detectors whose mAP is reported on the DIOR test set. The central claim is that MMO-IG generates realistic remote sensing images with accurate instance-level labels and that detectors pre-trained with these images perform better on real-world datasets.

Significance. If the claims hold, MMO-IG would be a practically useful tool for alleviating annotation scarcity in remote sensing object detection, because it produces both images and dense bounding-box labels in a controllable way. The idea of encoding class, location, and scale in a single iso-spacing grayscale map is simple and potentially reusable beyond this specific setting. The authors also release code, and the downstream evaluation across five detectors is a commendable effort. However, the current evidence is not yet convincing: several headline metrics are undefined, the SCDKG algorithm as written does not use the pid matrix, all results are single-run without statistical support, and the downstream gains are small and sometimes negative for individual classes. These issues prevent the paper from establishing its central claims as stated.

major comments (5)
  1. [Section III-B, Algorithm 1] The pid matrix is never actually used. In Algorithm 1, the line "class ← SAMPLE(pid matrix)" appears after the object has been appended to Lobj, and at the start of the next loop iteration the variable class is overwritten by "class ← SAMPLE(Pic)". Therefore the interdependence probability sampled from the pid matrix has no effect on any generated object's class, location, or scale. This contradicts the description in Section III-B and undermines the claimed pid-matrix improvements in Table II. Please correct the algorithm so that the previous class conditions the next class (or remove the pid matrix and rerun the ablation).
  2. [Section IV-A and Tables I-III] The CAS metric is used as a headline quality measure in Tables I, II, and III, and the paper also reports "CASzs" (zero-shot CAS), but CAS is never defined anywhere in the manuscript. The reader cannot tell what is measured, how it is computed, what its range is, or why higher values are better. Since the generation-quality claim rests substantially on CAS, this definition must be provided.
  3. [Section IV-D, Table IV] All downstream detection results are single-run numbers with no error bars, confidence intervals, or significance tests. The average per-class improvements from adding 20k generated images are small: +0.6 (R-CNN), +0.6 (Faster R-CNN), +0.4 (YOLO), +0.3 (PANet), and +0.1 (CornerNet) mAP, and several classes decline substantially (e.g., PANet Airplane drops 3.9 points, R-CNN Stadium drops 3.2 points). Without multiple seeds or a control condition (such as generated images with shuffled labels or an unrelated augmentation), the abstract's claim of "excellent performance" on real-world datasets is not established. Please provide statistical evidence and a control to show that the gains are attributable to the correctness of MMO-IG's instance-level labels rather than to generic effects of more training data.
  4. [Section IV-A, Tables I-II] The Accc and Accn evaluation protocol is not specified. It is unclear whether these numbers come from a trained detector, a frozen classifier, manual inspection, or a pixel-level comparison between ISIM and the generated image, and on how many images the evaluation is performed. Because the paper uses Accc and Accn to argue that ISIM regions decode to the intended classes and counts, the exact protocol must be described.
  5. [Section III-B and Section IV-B] The SCDKG priors (Pic, Pin, Psgc, and the pid matrix) are estimated from the DIOR dataset and then used to sample layouts for images whose FID and CAS are evaluated against DIOR. This makes part of the realism benchmark self-referential: the layout statistics are matched by construction, so the reported gains may overstate the improvement in image synthesis. Please clarify the reference set used for FID and CAS, and ideally evaluate on a held-out remote sensing dataset or compare against a baseline that uses real DIOR layout statistics without the full generative model. In addition, Table II as rendered is ambiguous about which components are active in each row, and the improvements attributed to the pid matrix in the text (stated as 3.8 in FIDzs and 1.1 in CAS on DIOR) do not match any adjacent row difference in the visible numbers; please clarify the ablation configuration and the exact comparisons.
minor comments (6)
  1. [Section IV-D] The sentence "we visualize the detection performance in Fig. 6" should refer to Fig. 10, which is the detection performance figure.
  2. [Section I] In the introduction, "generating RS images containing MMPs" appears to be a typo; the intended term is "MMOs".
  3. [Throughout] The notation for the probability distributions is inconsistent: the text alternates among "Pid", "pid", "𝒫", and "P" for the same quantities. Please standardize the notation.
  4. [Section III-C, Eq. (1)] Please clarify whether the class index m runs from 1 to M (with 0 reserved for background) and state explicitly that for M=20 the adjacent-class gray-level gaps are about 12 or 13 levels; this is directly relevant to the label-decoding accuracy discussion.
  5. [Section IV-A] The DIOR dataset reference is missing. The manuscript cites [54] for DIOR and DIOR-R, but [54] is an oriented object detection paper, not the original DIOR dataset paper. Please add the correct dataset citation.
  6. [Section IV-D] The sentence "although there is a decline in performance for a small number of object classes (averaging 15% per method)" is unclear and should be rephrased with the actual count or percentage of declining classes.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: MMO-IG is conditional image synthesis with independent downstream evaluation on the real DIOR test set.

full rationale

MMO-IG's derivation chain is: SCDKG statistics (Pic, Pin, Psgc, and the pid matrix) are estimated from DIOR annotations; layouts are sampled from these statistics; ISIM and SODI are rendered from those layouts; a diffusion model is trained to decode ISIM/SODI into images; realism is scored with FID/CAS against real DIOR imagery and controllability with Accc/Accn against the conditioning labels; and downstream detectors are trained on combined real and generated data and tested on the real DIOR test set (Table IV). Each stage is conditional on its inputs rather than assuming its outputs. In particular, the Accc and Accn metrics verify that the generated image actually follows the ISIM/SODI condition; these values could in principle be low, so they are not forced by construction. The FID/CAS evaluations use the same dataset that supplied the SCDKG priors, which is in-domain evaluation rather than a derivation that reduces to its input, and the downstream detection result on the held-out DIOR test set provides independent grounding. The authors' self-citations are to unrelated text-detection works and are not load-bearing. The absence of significance tests for the small downstream gains is a statistical-rigor concern, not circularity.

Assumptions & free parameters 4 free parameters · 4 assumptions · 3 invented entities

The pipeline is built on empirical statistics of DIOR: class frequencies, instance counts, per-class geometry distributions, and a co-occurrence matrix. These are fitted quantities, not first-principles constraints. The realism evaluation is entirely on the same dataset family, so the 'realistic and reliable' claims depend on FID/CAS and downstream transfer, both internal to DIOR.

free parameters (4)
  • Pic (instance class distribution) = Empirical class frequencies over 20 DIOR classes
    Used in Algorithm 1 to sample the class of each generated instance.
  • Pin (instance count distribution) = Empirical per-image instance count distribution from DIOR
    Used in Algorithm 1 to sample the number of instances per generated image.
  • Psgc (per-class spatial geometric characteristic distributions) = Empirical densities for location, scale, and aspect ratio per class from DIOR
    Used in Algorithm 1 to assign spatial attributes to each sampled instance.
  • pid matrix (interdependency probability matrix) = Empirical 20x20 conditional co-occurrence probabilities from DIOR
    Used in Algorithm 1 to sample each subsequent object class conditioned on the previous class.
assumptions (4)
  • domain assumption DIOR annotations are accurate and its class statistics are representative of real remote sensing imagery.
    SCDKG and all priors are fitted to DIOR; if the dataset statistics do not generalize, the generated layouts will be unrealistic. Stated in Section III-B and Section IV-A.
  • domain assumption The Stable Diffusion VAE and pretrained weights can be fine-tuned to decode ISIM regions into recognizable objects.
    The generative network is initialized with Stable Diffusion weights and trained with the Eq. (2) loss; the paper assumes this transfer works. Section III-E.
  • domain assumption CLIP text embeddings adequately capture the SODI count-based instruction.
    SODI is embedded via CLIP and used as a condition; the paper provides no analysis of how well CLIP distinguishes counts like '7 airplanes' from '8 airplanes'. Section III-E.
  • domain assumption FID and CAS are valid measures of generation quality for the RS detection use case.
    FID and CAS are used as the primary quality metrics, but CAS is not defined and no evidence is given that they correlate with downstream detection utility. Section IV-B.
invented entities (3)
  • ISIM (Iso-spacing Instance Map)
    purpose: Grayscale control signal encoding object class, location, and scale for the diffusion model.
    A design choice introduced by the paper; its effectiveness is only evaluated through the paper's own experiments and not independently validated.
  • SCDKG (Spatial-Cross Dependency Knowledge Graph)
    purpose: Models class interdependencies and spatial geometric characteristics for sampling realistic object layouts.
    An empirical graph fitted to DIOR statistics; it is not an independently evidenced structure and its realism claims are only tested on DIOR-style images.
  • SODI (Structured Object Distribution Instruction)
    purpose: Text prompt that lists object class counts to keep generated images aligned with ISIM labels.
    A text-template method introduced by the paper; its benefit is shown only through the paper's ablations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MMO-IG: Multi-Class and Multi-Scale Object Image Generation for Remote Sensing." pith.science (2026). https://pith.science/paper/H43M6XDO

@misc{pith2026241213684,
  author       = {Pith},
  title        = {Pith review of: MMO-IG: Multi-Class and Multi-Scale Object Image Generation for Remote Sensing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/H43M6XDO}},
  note         = {Machine review of arXiv:2412.13684}
}
read the original abstract

The rapid advancement of deep generative models (DGMs) has significantly advanced research in computer vision, providing a cost-effective alternative to acquiring vast quantities of expensive imagery. However, existing methods predominantly focus on synthesizing remote sensing (RS) images aligned with real images in a global layout view, which limits their applicability in RS image object detection (RSIOD) research. To address these challenges, we propose a multi-class and multi-scale object image generator based on DGMs, termed MMO-IG, designed to generate RS images with supervised object labels from global and local aspects simultaneously. Specifically, from the local view, MMO-IG encodes various RS instances using an iso-spacing instance map (ISIM). During the generation process, it decodes each instance region with iso-spacing value in ISIM-corresponding to both background and foreground instances-to produce RS images through the denoising process of diffusion models. Considering the complex interdependencies among MMOs, we construct a spatial-cross dependency knowledge graph (SCDKG). This ensures a realistic and reliable multidirectional distribution among MMOs for region embedding, thereby reducing the discrepancy between source and target domains. Besides, we propose a structured object distribution instruction (SODI) to guide the generation of synthesized RS image content from a global aspect with SCDKG-based ISIM together. Extensive experimental results demonstrate that our MMO-IG exhibits superior generation capabilities for RS images with dense MMO-supervised labels, and RS detectors pre-trained with MMO-IG show excellent performance on real-world datasets.

Figures

Figures reproduced from arXiv: 2412.13684 by the authors.

Figure 1
Figure 1. Illustration of the generation of RS images containing MMOs by the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overall pipeline of MMO-IG for generating RS images with dense instance-level bounding box labels. It first synthesizes rational spatial geometric [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Illustration of the proposed SCDKG, which models complex inter [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Illustration of the proposed ISIM encodes instances with different [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Illustration of the proposed SODI generation process. It consists of [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Visualization comparison of the generated image by our MMO-IG with different text conditions on the DIOR dataset. Considering the grayscale values [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 8
Figure 8. Figure 8: Visualization of some samples with irrational aspect ratio, scale, and [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Visualization of some qualified generated RS images by our MMO-IG. Considering the grayscale values of some objects are too low and difficult to [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 10
Figure 10. Figure 10: Comparison of different detectors’ accuracy across each category on the DIOR dataset under the setting of data augmentation with 20k generation [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 42 canonical work pages

  1. [2]

    Efficient inductive vision transformer for oriented object detection in remote sensing imagery,

    C. Zhang, J. Su, Y . Ju, K.-M. Lam, and Q. Wang, “Efficient inductive vision transformer for oriented object detection in remote sensing imagery,” IEEE Transactions on Geoscience and Remote Sensing , 2023

  2. [3]

    Global to local: A scale-aware network for remote sensing object detection,

    T. Gao, Q. Niu, J. Zhang, T. Chen, S. Mei, and A. Jubair, “Global to local: A scale-aware network for remote sensing object detection,” IEEE Transactions on Geoscience and Remote Sensing , 2023

  3. [4]

    Large selec- tive kernel network for remote sensing object detection,

    Y . Li, Q. Hou, Z. Zheng, M.-M. Cheng, J. Yang, and X. Li, “Large selec- tive kernel network for remote sensing object detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 16 794–16 805

  4. [5]

    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

  5. [6]

    Cm-net: Concen- tric mask based arbitrary-shaped text detection,

    C. Yang, M. Chen, Z. Xiong, Y . Yuan, and Q. Wang, “Cm-net: Concen- tric mask based arbitrary-shaped text detection,” IEEE Transactions on Image Processing, vol. 31, pp. 2864–2877, 2022

  6. [7]

    Rich feature hierarchies for accurate object detection and semantic segmentation,

    R. Girshick, J. Donahue, T. Darrell, and J. Malik, “Rich feature hierarchies for accurate object detection and semantic segmentation,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2014, pp. 580–587

  7. [8]

    Text growing on leaf,

    C. Yang, M. Chen, Y . Yuan, and Q. Wang, “Text growing on leaf,”IEEE Transactions on Multimedia , vol. 25, pp. 9029–9043, 2023

  8. [9]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems , vol. 33, pp. 6840– 6851, 2020

Show all 54 references
  1. [10]

    High- resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High- resolution image synthesis with latent diffusion models,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2022, pp. 10 684–10 695

  2. [11]

    Adding conditional control to text-to-image diffusion models,

    L. Zhang, A. Rao, and M. Agrawala, “Adding conditional control to text-to-image diffusion models,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 3836–3847

  3. [12]

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

    D. Tang, X. Cao, X. Hou, Z. Jiang, J. Liu, and D. Meng, “Crs-diff: Controllable remote sensing image generation with diffusion model,” IEEE Transactions on Geoscience and Remote Sensing , 2024

  4. [13]

    Generating and sifting pseudolabeled samples for improving the per- formance of remote sensing image scene classification,

    X. Qian, X. Chen, W. Yue, X. Liu, J. Guo, Z. Li, Y . Li, and W. Wang, “Generating and sifting pseudolabeled samples for improving the per- formance of remote sensing image scene classification,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing ,...

  5. [14]

    Gligen: Open-set grounded text-to-image generation,

    Y . Li, H. Liu, Q. Wu, F. Mu, J. Yang, J. Gao, C. Li, and Y . J. Lee, “Gligen: Open-set grounded text-to-image generation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 22 511–22 521

  6. [15]

    Layoutdiffusion: Controllable diffusion model for layout-to-image generation,

    G. Zheng, X. Zhou, X. Li, Z. Qi, Y . Shan, and X. Li, “Layoutdiffusion: Controllable diffusion model for layout-to-image generation,” in Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 22 490–22 499

  7. [16]

    Reco: Region-controlled text-to-image generation,

    Z. Yang, J. Wang, Z. Gan, L. Li, K. Lin, C. Wu, N. Duan, Z. Liu, C. Liu, M. Zeng et al. , “Reco: Region-controlled text-to-image generation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 14 246–14 255

  8. [17]

    Image synthesis from reconfigurable layout and style,

    W. Sun and T. Wu, “Image synthesis from reconfigurable layout and style,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 10 531–10 540

  9. [18]

    Uni-controlnet: All-in-one control to text-to-image diffusion models,

    S. Zhao, D. Chen, Y .-C. Chen, J. Bao, S. Hao, L. Yuan, and K.-Y . K. Wong, “Uni-controlnet: All-in-one control to text-to-image diffusion models,” Advances in Neural Information Processing Systems , vol. 36, 2024

  10. [19]

    Controlnet++: Improving conditional controls with efficient consistency feedback,

    M. Li, T. Yang, H. Kuang, J. Wu, Z. Wang, X. Xiao, and C. Chen, “Controlnet++: Improving conditional controls with efficient consistency feedback,” in European Conference on Computer Vision . Springer, 2025, pp. 129–147

  11. [20]

    Controlnet-xs: Designing an efficient and effective architecture for controlling text-to-image diffusion models,

    D. Zavadski, J.-F. Feiden, and C. Rother, “Controlnet-xs: Designing an efficient and effective architecture for controlling text-to-image diffusion models,” arXiv preprint arXiv:2312.06573 , 2023

  12. [21]

    Yolov3: An incremental improvement,

    J. Redmon, “Yolov3: An incremental improvement,” arXiv preprint arXiv:1804.02767, 2018

  13. [22]

    Bip-net: Bidirectional perspective strategy based arbitrary-shaped text detection network,

    C. Yang, M. Chen, Y . Yuan, and Q. Wang, “Bip-net: Bidirectional perspective strategy based arbitrary-shaped text detection network,” in ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2022, pp. 2255–2259

  14. [23]

    Path aggregation network for instance segmentation,

    S. Liu, L. Qi, H. Qin, J. Shi, and J. Jia, “Path aggregation network for instance segmentation,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 8759–8768

  15. [24]

    Corporate event predictions using large language models,

    Z. Xiao, Z. Mai, Z. Xu, Y . Cui, and J. Li, “Corporate event predictions using large language models,” in 2023 10th International Conference on Soft Computing & Machine Intelligence (ISCMI) . IEEE, 2023, pp. 193–197

  16. [25]

    Cornernet: Detecting objects as paired keypoints,

    H. Law and J. Deng, “Cornernet: Detecting objects as paired keypoints,” in Proceedings of the European conference on computer vision (ECCV) , 2018, pp. 734–750

  17. [26]

    Instance mask growing on leaf

    C. Yang, H. Ma, and Q. Wang, “Instance mask growing on leaf.” in BMVC, 2023, pp. 4–6

  18. [27]

    Cross-layer attention network for small object detection in remote sensing imagery,

    Y . Li, Q. Huang, X. Pei, Y . Chen, L. Jiao, and R. Shang, “Cross-layer attention network for small object detection in remote sensing imagery,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , vol. 14, pp. 2148–2161, 2020

  19. [28]

    Superyolo: Super resolution assisted object detection in multimodal remote sensing im- agery,

    J. Zhang, J. Lei, W. Xie, Z. Fang, Y . Li, and Q. Du, “Superyolo: Super resolution assisted object detection in multimodal remote sensing im- agery,” IEEE Transactions on Geoscience and Remote Sensing , vol. 61, pp. 1–15, 2023

  20. [29]

    Building a bridge of bounding box regression between oriented and horizontal object detection in remote sensing images,

    X. Qian, B. Wu, G. Cheng, X. Yao, W. Wang, and J. Han, “Building a bridge of bounding box regression between oriented and horizontal object detection in remote sensing images,” IEEE Transactions on Geoscience and Remote Sensing , vol. 61, pp. 1–9, 2023

  21. [30]

    Feature alignment fpn for oriented object detection in remote sensing images,

    Z. Li, E. Li, T. Xu, A. Samat, and W. Liu, “Feature alignment fpn for oriented object detection in remote sensing images,” IEEE Geoscience and Remote Sensing Letters , vol. 20, pp. 1–5, 2023

  22. [31]

    Tbnet: A texture and boundary-aware network for small weak object detection in remote- sensing imagery,

    Z. Li, Y . Wang, D. Xu, Y . Gao, and T. Zhao, “Tbnet: A texture and boundary-aware network for small weak object detection in remote- sensing imagery,” Pattern Recognition, vol. 158, p. 110976, 2025

  23. [32]

    Small object detection in remote sensing images based on super-resolution,

    F. Xiaolin, H. Fan, Y . Ming, Z. Tongxin, B. Ran, Z. Zenghui, and G. Zhiyuan, “Small object detection in remote sensing images based on super-resolution,” Pattern Recognition Letters , vol. 153, pp. 107–112, 2022

  24. [33]

    Feature split–merge–enhancement network for remote sensing object detection,

    W. Ma, N. Li, H. Zhu, L. Jiao, X. Tang, Y . Guo, and B. Hou, “Feature split–merge–enhancement network for remote sensing object detection,” IEEE Transactions on Geoscience and Remote Sensing , vol. 60, pp. 1– 17, 2022

  25. [34]

    An adaptive attention fusion mechanism convolutional network for object detection in remote sensing images,

    Y . Ye, X. Ren, B. Zhu, T. Tang, X. Tan, Y . Gui, and Q. Yao, “An adaptive attention fusion mechanism convolutional network for object detection in remote sensing images,” Remote Sensing, vol. 14, no. 3, p. 516, 2022

  26. [35]

    Attention-based multi-level feature fusion for object detection in remote sensing images,

    X. Dong, Y . Qin, Y . Gao, R. Fu, S. Liu, and Y . Ye, “Attention-based multi-level feature fusion for object detection in remote sensing images,” Remote Sensing , vol. 14, no. 15, p. 3735, 2022

  27. [36]

    Swin-transformer-enabled yolov5 with atten- tion mechanism for small object detection on satellite images,

    H. Gong, T. Mu, Q. Li, H. Dai, C. Li, Z. He, W. Wang, F. Han, A. Tuniyazi, H. Li et al., “Swin-transformer-enabled yolov5 with atten- tion mechanism for small object detection on satellite images,” Remote Sensing, vol. 14, no. 12, p. 2861, 2022

  28. [37]

    Rotation- aware and multi-scale convolutional neural network for object detection in remote sensing images,

    K. Fu, Z. Chang, Y . Zhang, G. Xu, K. Zhang, and X. Sun, “Rotation- aware and multi-scale convolutional neural network for object detection in remote sensing images,” ISPRS Journal of Photogrammetry and Remote Sensing , vol. 161, pp. 294–308, 2020

  29. [38]

    Dense label encoding for boundary discontinuity free rotation detection,

    X. Yang, L. Hou, Y . Zhou, W. Wang, and J. Yan, “Dense label encoding for boundary discontinuity free rotation detection,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2021, pp. 15 819–15 829

  30. [39]

    Instance-aware distillation for efficient object detection in remote sensing images,

    C. Li, G. Cheng, G. Wang, P. Zhou, and J. Han, “Instance-aware distillation for efficient object detection in remote sensing images,” IEEE Transactions on Geoscience and Remote Sensing , vol. 61, pp. 1–11, 2023

  31. [40]

    YOLO-DA: an efficient yolo- based detector for remote sensing object detection,

    J. Lin, Y . Zhao, S. Wang, and Y . Tang, “YOLO-DA: an efficient yolo- based detector for remote sensing object detection,” IEEE Geoscience and Remote Sensing Letters , vol. 20, pp. 1–5, 2023

  32. [41]

    Fast tiny object detection in large-scale remote sensing images,

    J. Pang, C. Li, J. Shi, Z. Xu, and H. Feng, “Fast tiny object detection in large-scale remote sensing images,” IEEE Transactions on Geoscience and Remote Sensing , vol. 57, no. 8, pp. 5512–5524, 2019

  33. [42]

    Auto-encoding variational bayes,

    D. P. Kingma and M. Welling, “Auto-encoding variational bayes,” in 2nd International Conference on Learning Representations, ICLR 2014, 12 Banff, AB, Canada, April 14-16, 2014, Conference Track Proceedings , Y . Bengio and Y . LeCun, Eds., 2014

  34. [43]

    Vari- ational autoencoder for deep learning of images, labels and captions,

    Y . Pu, Z. Gan, R. Henao, X. Yuan, C. Li, A. Stevens, and L. Carin, “Vari- ational autoencoder for deep learning of images, labels and captions,” in Advances in Neural Information Processing Systems 29: Annual Conference on Neural Information Processing Systems 2016, December ...

  35. [44]

    Grammar varia- tional autoencoder,

    M. J. Kusner, B. Paige, and J. M. Hern ´andez-Lobato, “Grammar varia- tional autoencoder,” in Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW, Australia, 6-11 August 2017, ser. Proceedings of Machine Learning Research, D. Precup and...

  36. [45]

    Generative adversarial nets,

    I. J. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. C. Courville, and Y . Bengio, “Generative adversarial nets,” in Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014, December 8-...

  37. [46]

    Coupled generative adversarial networks,

    M. Liu and O. Tuzel, “Coupled generative adversarial networks,” in Advances in Neural Information Processing Systems 29: Annual Con- ference on Neural Information Processing Systems 2016, December 5- 10, 2016, Barcelona, Spain , D. D. Lee, M. Sugiyama, U. von Luxburg, I. Guyon...

  38. [47]

    Least squares generative adversarial networks,

    X. Mao, Q. Li, H. Xie, R. Y . K. Lau, Z. Wang, and S. P. Smolley, “Least squares generative adversarial networks,” in IEEE International Conference on Computer Vision, ICCV 2017, V enice, Italy, October 22- 29, 2017 . IEEE Computer Society, 2017, pp. 2813–2821

  39. [48]

    Deep unsupervised learning using nonequilibrium thermodynamics,

    J. Sohl-Dickstein, E. Weiss, N. Maheswaranathan, and S. Ganguli, “Deep unsupervised learning using nonequilibrium thermodynamics,” in International conference on machine learning . PMLR, 2015, pp. 2256–2265

  40. [49]

    Singan: Learning a generative model from a single natural image,

    T. R. Shaham, T. Dekel, and T. Michaeli, “Singan: Learning a generative model from a single natural image,” in Proceedings of the IEEE/CVF international conference on computer vision , 2019, pp. 4570–4580

  41. [50]

    Remote sensing data augmentation through adversarial training,

    N. Lv, H. Ma, C. Chen, Q. Pei, Y . Zhou, F. Xiao, and J. Li, “Remote sensing data augmentation through adversarial training,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , vol. 14, pp. 9318–9333, 2021

  42. [51]

    Rsdiff: Remote sensing image generation from text using diffusion model,

    A. Sebaq and M. ElHelw, “Rsdiff: Remote sensing image generation from text using diffusion model,” Neural Computing and Applications , pp. 1–9, 2024

  43. [52]

    Learning transferable visual models from natural language supervi- sion,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark, G. Krueger, and I. Sutskever, “Learning transferable visual models from natural language supervi- sion,” in Proceedings of the 38th International Conference on Machine...

  44. [53]

    Taming transformers for high- resolution image synthesis,

    P. Esser, R. Rombach, and B. Ommer, “Taming transformers for high- resolution image synthesis,” in IEEE Conference on Computer Vision and Pattern Recognition 2021, virtual, June 19-25, 2021 , 2021, pp. 12 873–12 883

  45. [54]

    Anchor- free oriented proposal generator for object detection,

    G. Cheng, J. Wang, K. Li, X. Xie, C. Lang, Y . Yao, and J. Han, “Anchor- free oriented proposal generator for object detection,” IEEE Transactions on Geoscience and Remote Sensing , vol. 60, pp. 1–11, 2022

  46. [55]

    Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation,

    J. Li, D. Li, C. Xiong, and S. Hoi, “Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation,” in International conference on machine learning . PMLR, 2022, pp. 12 888–12 900. Chuang Yang received the B.E. degree in automa- tion...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.