Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

Cross-DINO: Cross the Deep MLP and Transformer for Small Object Detection

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read Cross-DINO claims that three coordinated fixes — a context-rich MLP backbone, a twice-gated encoder fusion, and a size-aware classification loss — raise COCO small-object AP_S from 32.0 to 36.4 while using fewer parameters than the DINO…

desk verdict The +4.4 AP_S claim is probably real, but the Boost loss doesn't do what the paper says it does, and the beta selection is internally inconsistent. read the letter →

arxiv 2505.21868 v1 pith:JDQNKYIN submitted 2025-05-28 cs.CV

classification cs.CV
keywords smallobjectdetectionDETR-likedetectorsdeepMLPbackboneCrossCodingTwiceModuleBoostlossCategory-Sizesoftlabeltransformerencoder
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

Cross-DINO is a proposal for making DETR-style transformer detectors substantially better at finding small objects. The paper argues that three structural weaknesses — context-poor backbone features, blur introduced by repeated encoder attention, and classification confidence that drops as objects shrink — can each be fixed, and that fixing all three compounds. Its recipe replaces or augments the CNN backbone with a deep MLP that captures short- and long-range context in one pass, feeds fine-grained backbone detail back into the encoder through a twice-gated Cross Coding Twice Module, and reweights the classification loss with a new soft label that mixes class identity with object size. On COCO val2017, the combined model reaches 36.4% $AP_S$ with 45M parameters under a 12-epoch schedule, 4.4 points above the DINO baseline, and the gains carry over to WiderPerson, VisDrone2019, and SODA-D.

What carries the argument

The load-bearing mechanism is the Category-Size (CS) soft label, $cs_i = \sqrt{(h_i/H)(w_i/W)}\, y_i$, which fuses each object's class label with the square root of its area ratio to the image, and the Boost loss built on it; this is the component that directly targets low class-prediction scores for small objects. Two supporting mechanisms carry the rest of the argument: the CLAP-Strip-MLP backbone, an attention-free MLP whose overlapping-patch cropping makes it resolution-adaptive while mixing short- and long-range context, and the Cross Coding Twice Module (CCTM), a twice-gated fusion that re-injects fine-grained backbone detail into the blur-prone transformer encoder features. Together they supply richer initial features, counteract encoder blur, and reweight classification loss by object size.

What would settle it

On COCO val2017 with the same 12-epoch DINO recipe, replace the CS target $cs_i = \sqrt{(h_i/H)(w_i/W)}\, y_i$ with $cs_i = ((h_i/H)(w_i/W))^p\, y_i$ for several exponents p (e.g., 0.25, 0.5, 1, 2) and for a size-only reweighting with no class term; if AP_S stays roughly flat across p and across the no-class variant, the specific square-root category-size form is not what carries the gain.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the small-object deficit of DETR-like detectors is not one problem but three, and that a detector built to address all three at once beats the standard DINO baseline by +4.4% $AP_S$ (36.4% vs. 32.0%) on COCO val2017 using fewer parameters (45M vs. 47M) under the same 12-epoch training. The first fix is a deep MLP backbone, CLAP-Strip-MLP, whose strip-wise token mixing captures both local and global context; a cropping-with-overlap scheme lets this fixed-resolution MLP accept arbitrary image sizes. The second is CCTM, which uses two gated crossings to let transformer encoder features absorb fine-grained detail from backbone features, countering the blur that repeated attention layers impose on small objects. The third is Boost loss, whose Category-Size soft label $cs_i = \sqrt{(h_i/H)(w_i/W)}\, y_i$ reweights each positive classification term by object size, raising the training pressure on exactly the small objects whose confidence scores lag. The paper reports consistent $AP_S$ gains across ResNet50, Swin-T, and CLAP-Strip-T backbones, and improvements over DINO on WiderPerson, VisDrone2019, and SODA-D, while noting that on AI-TOD the CLAP-Strip-T backbone's fewer channels hurt until the input resolution is increased.

Load-bearing premise

The recipe depends on the assumption that multiplying a box's class label by the square root of its area ratio to the image, used as the training target, is what improves small-object detection, rather than just any stronger emphasis on small boxes.

Editorial extensions

If this is right

  • A DETR-like detector can gain 4.4 points of small-object AP on COCO val2017 without adding parameters, since Cross-DINO with CLAP-Strip-T runs at 45M parameters versus DINO's 47M.
  • The CLAP cropping scheme converts fixed-resolution deep MLP backbones into general-purpose, arbitrary-resolution backbones, making them usable in dense prediction tasks beyond classification.
  • Injecting backbone detail into encoder features through twice-gated fusion is a workable antidote to encoder feature blur for small objects.
  • Boost loss raises classification attention on small objects: on all three backbones tested, replacing the standard focal loss with Boost loss improves AP_S, with the largest single-component gain on CLAP-Strip-T (+1.2 AP_S).
  • The gains transfer across scene types: +3.7 AP on VisDrone2019, +4.3 AP on SODA-D, and +0.86 AP on WiderPerson with the Swin-T backbone.

Reading between the lines

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

  • Editorial extension: the CS soft label is a geometric reweighting of standard classification loss; a natural test is whether the square-root area ratio is special or whether any monotone size function (e.g., linear or log area) gives the same gain.
  • Editorial extension: CCTM's twice-gated fusion could be dropped into other encoder-decoder dense-prediction models, such as segmentation or keypoint detectors, wherever encoder downsampling blurs small structures.
  • Editorial extension: the AI-TOD result suggests a boundary condition — an MLP backbone with 31.25% of ResNet's channels underserves very tiny objects — so the paper's recipe may need a channel-wider MLP or higher input resolution to keep its edge in extreme tiny-object regimes.
  • Editorial extension: because the paper does not isolate feature-resolution from context-aggregation effects, an alternative explanation for part of the gain is that the deeper MLP simply provides a higher-quality feature pyramid; comparing Cross-DINO against DINO with a high-resolution ResNet variant would separate the two.
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

4 major / 6 minor

Summary. Cross-DINO extends the DINO DETR-style detector with three components aimed at small object detection: (1) CLAP-Strip-T, an adaptation of the Strip-MLP backbone to arbitrary input resolutions via overlapping crops; (2) a Cross Coding Twice Module (CCTM) that fuses backbone features with transformer encoder features; and (3) a Boost Loss with a new 'Category-Size' (CS) soft label that is meant to re-weight classification objective toward small objects. The paper reports consistent gains over DINO on COCO val2017 (36.4% AP_S vs. 32.0% with the CLAP-Strip-T backbone), WiderPerson, VisDrone2019, SODA-D, and smaller gains on AI-TOD, using a 12-epoch training schedule and comparable parameter counts.

Significance. If the claims hold, the paper would provide a useful practical recipe for improving small-object detection in DETR-like detectors, and it would demonstrate that MLP-style backbones can be adapted to dense prediction tasks without retraining. The paper includes a broad set of comparisons and ablations across five datasets, which is a strength. However, the central novelty that is most load-bearing—the Boost Loss and its CS soft-label formulation—is currently not adequately justified quantitatively, and the stated mechanism does not match the equation as written. The backbone and module contributions are plausible and well evaluated, but the loss contribution needs correction or re-analysis before the headline result can be fully credited.

major comments (4)
  1. [Sec. III-D, Eq. (5)] The positive term of the Boost loss is α(1 - ĉs^β)^γ cs^β log(p_i). Since cs_i = sqrt((h_i/H)(w_i/W)) decreases with object size, the multiplier cs^β also decreases with decreasing size (for β=1, a 8×8 object in a 640×640 image gives cs≈0.0125 while an 80×80 object gives cs≈0.096). The claimed effect that 'smaller objects get larger loss' is therefore not implemented by this equation; the product (1 - ĉs^β)^γ cs^β is not monotone in size and can be inverted over plausible ranges. Table IX in the appendix computes the relative distance only for the (1 - ĉs^β)^γ factor, omitting the cs^β multiplier, so it does not support the stated behavior. Please provide a corrected loss form, a numerical demonstration that the full weight is monotone in the intended direction, or an explicit statement that the implemented loss is a different form than Eq. (5).
  2. [Sec. III-D, Eq. (4)-(5)] The paper repeatedly calls CS a 'soft label' and says it is used as 'new ground truth', but Eq. (5) is not cross-entropy with respect to the soft label cs. A standard soft-label cross-entropy for binary classification would include -(1-cs)log(1-p) in the positive-class term; here cs appears only as a multiplicative weight in the term -(cs^β)log(p). This is a semantic mismatch that affects the interpretation of the method. Please clarify whether cs is a soft label or a loss weight, and if it is a weight, justify the square-root-area functional form with a comparison to other size transforms (e.g., linear area, log-area, or relative-area-to-image).
  3. [Table VI and Table X] The hyperparameter β is set differently across settings without a sensitivity study on COCO: β=1.0 for COCO-ResNet50, β=0.1 for COCO-Swin-T, β=0.1 for SODA-D, β=0.05 for AI-TOD. Table X, the only β ablation, is on VisDrone and shows that β=0.25 gives a higher AP (35.8) than the chosen β=0.1 (35.4), with only a small drop in AP_vt; the paper does not explain why β=0.1 is then preferred. Because Boost Loss is credited with gains up to +1.2 AP_S in the CLAP-Strip-T ablation (Table VI), the dependence of the method on a per-dataset, per-backbone β choice is load-bearing and must be analyzed on the main benchmark (COCO) and, ideally, with a principled selection rule.
  4. [Sec. III-B and Appendix 'Methodology of CLAP'] The CLAP method is essential for using Strip-MLP as a general backbone, but it is only validated through an ERF visualization (Fig. 11) and indirect detection results. No ablation compares CLAP with alternative resolution-adaptation strategies (e.g., tiling without overlap, padding to a multiple of the fixed size, or interpolation to the fixed size), and no sensitivity to the overlap sizes lw/lh is reported. A quantitative comparison or an ablation on the overlap parameter would significantly strengthen the claim that the proposed cropping scheme, rather than the backbone capacity difference, is responsible for the observed gains.
minor comments (6)
  1. [Eq. (4)-(5), notation] The notation for y_i is ambiguous: y_i could be a scalar class index or a one-hot vector, and the loss in Eq. (5) is written in a form that resembles binary classification. Please define the multi-class version (e.g., with sum over classes) and state how p_i is computed from the model's softmax.
  2. [Fig. 3 and Sec. III-D] The caption of Fig. 3 says that Cross-DINO detects 'harder' small objects with lower scores, resulting in a decrease in average confidence scores, which appears to contradict the motivation of Boost Loss as improving class prediction scores. The main text should discuss this trade-off explicitly.
  3. [Table I] The row 'DINO [49] Swin-T' cites reference [49] (detrex) rather than the original DINO paper; this should be clarified so readers know the baseline source.
  4. [Sec. III-C, Eq. (3)] The fixed weight of 2 for the encoder feature in Ecf = 2E·B'E' + B·(1-B'E') is introduced without an ablation or sensitivity analysis; a brief justification or a small study would make this design choice transparent.
  5. [Appendix, Table IX] In the caption of Table IX, γ is listed as 0.25, while the main text states the default γ=2.0; the inconsistency is confusing and should be corrected.
  6. [Sec. I and Conclusion] The paper says 'we introduce a new deep MLP model' but actually the deep MLP component is the existing Strip-MLP adapted by CLAP; rephrase to avoid overclaiming novelty of the backbone architecture itself.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: external benchmarks and independent ablations ground the claims.

full rationale

Cross-DINO's central claims are empirical results on external benchmarks (COCO val2017, VisDrone2019 test-dev, SODA-D test, WiderPerson, AI-TOD), not quantities derived by definition from the method's own equations. The CLAP-Strip-T backbone is built on the authors' earlier Strip-MLP (ref. [5]), but that citation is an independently published ICCV paper with released code, and its contribution here is established by the paper's own DINO-CLAP-Strip-T ablation (+3.0 AP_S over DINO), not by an appeal to authority or a uniqueness theorem. CCTM is a gating module whose internal design is tested by component ablations (Table VIII), and Boost loss is a loss-function proposal whose effect is measured by train/val comparisons (Tables VI, XI). The appendix does reveal that beta was selected on the VisDrone2019 test set (Table X), which is a legitimate test-set-tuning concern and a correctness risk, but it is not circularity: the reported AP is not forced by construction from the definition of cs, and no equation in the paper reduces a predicted quantity to a fitted parameter or defines a component in terms of the target metric. The skeptic's concern that Eq. (5)'s cs^beta multiplier may not implement the stated 'smaller objects get larger loss' mechanism is a mathematical-correctness issue, not circularity, since the loss's effectiveness is an empirical finding rather than an input assumption. Self-citations [4] and [5] are not load-bearing as arguments. Therefore no circular step is exhibited.

Assumptions & free parameters 4 free parameters · 5 assumptions · 2 invented entities

The paper's central gains come from the backbone, the fusion module, and the loss. The free parameters alpha, gamma, beta, and the CCTM weight 2 are set by hand or by limited ablations, with beta varying per dataset. The CLAP technique and CS soft label are new entities whose validity rests on this paper's own experiments, so they are listed as invented with no independent evidence. The axioms are mostly standard deep-learning assumptions plus the specific functional forms of the new modules.

free parameters (4)
  • alpha (α) = 0.25
    Hyperparameter in Boost loss (Eq. 5) balancing positive and negative terms. Set by default to 0.25 with no ablation study shown for α.
  • gamma (γ) = 2.0
    Hyperparameter in Boost loss (Eq. 5) controlling the scaling factor exponent. Set by default to 2.0 with no ablation study shown for γ.
  • beta (β) = 1.0 for COCO, 0.1 for VisDrone/SODA-D, 0.05 for AI-TOD
    Scaling factor for cs in Boost loss. Chosen per dataset to balance object-size weight distributions. The appendix shows a beta ablation on VisDrone only, where beta=0.25 gives higher AP than beta=0.1, but beta=0.1 is used for the main result; no COCO or SODA-D beta ablation is shown.
  • CCTM encoder-feature weight of 2 = 2
    In Eq. 3, Ecf = 2E·B'E' + B·(1-B'E'). The weight 2 is stated as heuristic to emphasize encoder features; no ablation on this weight is provided.
assumptions (5)
  • domain assumption The soft label cs = sqrt((h/H)*(w/W))*y correctly encodes the desired relationship between object size and classification confidence.
    Invoked in Eq. 4 in Sec. III-D. The paper motivates that small objects have low class scores, but provides no derivation or comparative study showing this particular functional form is the right one.
  • domain assumption The backbone feature B and encoder feature E are compatible for element-wise gating in CCTM because they have the same spatial resolution and feature dimension.
    Assumed in Sec. III-C, Eqs. 1-3. The paper does not discuss how the 4-scale features are aligned or why direct element-wise gating is safe across different feature semantics.
  • ad hoc to paper Applying the pre-trained Strip-MLP as overlapping patches using CLAP preserves the information aggregation properties of the original fixed-resolution model.
    Introduced in Sec. III-B and Appendix. Patch overlap and averaging with weight 0.5 is a heuristic. The paper provides ERF visualizations but no quantitative analysis of how CLAP changes token interactions versus native resolution processing.
  • domain assumption DINO's training recipe and post-processing are sufficient for a fair comparison, and the models are not overfitting to small-object metrics despite no error bars.
    All experiments follow DINO or cited datasets' recipes, with a single run per setting. The paper does not report variance or multiple seeds.
  • domain assumption The AP and related metrics on COCO, VisDrone, AI-TOD, SODA-D, and WiderPerson are the right evaluation for SOD and improvements on them are not metric-specific artifacts.
    Standard in the field; the paper uses them without discussing potential matching biases (e.g., small-object ground-truth quality). Not a flaw per se, but an unstated background premise.
invented entities (2)
  • CLAP method (Cropping with Adaptive overLaPping)
    purpose: Enables fixed-resolution MLP backbones to process arbitrary-size images by cropping into overlapping patches with shared weights.
    The method is only validated within this paper's experiments and ERF visualizations. It is a technique, not a falsifiable entity with a standalone prediction.
  • Category-Size (CS) soft label
    purpose: Integrates category and object size into the classification target to increase loss weight for small objects.
    A new labeling quantity defined in Eq. 4. Its validity is supported only by the loss results in this paper; no independent theoretical or empirical validation is provided outside this work.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Cross-DINO: Cross the Deep MLP and Transformer for Small Object Detection." pith.science (2026). https://pith.science/paper/JDQNKYIN

@misc{pith2026250521868,
  author       = {Pith},
  title        = {Pith review of: Cross-DINO: Cross the Deep MLP and Transformer for Small Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JDQNKYIN}},
  note         = {Machine review of arXiv:2505.21868}
}
read the original abstract

Small Object Detection (SOD) poses significant challenges due to limited information and the model's low class prediction score. While Transformer-based detectors have shown promising performance, their potential for SOD remains largely unexplored. In typical DETR-like frameworks, the CNN backbone network, specialized in aggregating local information, struggles to capture the necessary contextual information for SOD. The multiple attention layers in the Transformer Encoder face difficulties in effectively attending to small objects and can also lead to blurring of features. Furthermore, the model's lower class prediction score of small objects compared to large objects further increases the difficulty of SOD. To address these challenges, we introduce a novel approach called Cross-DINO. This approach incorporates the deep MLP network to aggregate initial feature representations with both short and long range information for SOD. Then, a new Cross Coding Twice Module (CCTM) is applied to integrate these initial representations to the Transformer Encoder feature, enhancing the details of small objects. Additionally, we introduce a new kind of soft label named Category-Size (CS), integrating the Category and Size of objects. By treating CS as new ground truth, we propose a new loss function called Boost Loss to improve the class prediction score of the model. Extensive experimental results on COCO, WiderPerson, VisDrone, AI-TOD, and SODA-D datasets demonstrate that Cross-DINO efficiently improves the performance of DETR-like models on SOD. Specifically, our model achieves 36.4% APs on COCO for SOD with only 45M parameters, outperforming the DINO by +4.4% APS (36.4% vs. 32.0%) with fewer parameters and FLOPs, under 12 epochs training setting. The source codes will be available at https://github.com/Med-Process/Cross-DINO.

Figures

Figures reproduced from arXiv: 2505.21868 by the authors.

Figure 1
Figure 1. Comparison of different models of AP and AP [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Challenges of SOD. (a) In contrast to large objects, small objects [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 4
Figure 4. The overall architecture of the proposed Cross-DINO. Our Cross-DINO architecture utilizes a 4-scale backbone feature maps for the decoder to [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (7 more)
Figure 5
Figure 5. Figure 5: The architecture of Cross Coding Twice Module. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Comparison of the features from different parts of the Cross-DINO model. The visualization shows one channel of these features. The backbone [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: The performance curves for DINO-R50 and Cross-DINO (abbreviated [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Solution to accept input images with arbitrary sizes for deep MLP [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: The visualization of detection results of DINO-ResNet50 [68] and DINO-ResNet50 with our CCTM and Boost Loss (noted as DINO-CB) on [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]
Figure 10
Figure 10. Figure 10: The visualization of detection results of DINO [68] and Cross-DINO on validation images from the WiderPerson [70] dataset. The visualization [PITH_FULL_IMAGE:figures/full_fig_p015_10.png]
Figure 11
Figure 11. Figure 11: Visualization of Effective Receptive Fields (ERF) between our CLAP-Strip-T and other benchmark models. Pixels with higher intensity signify [PITH_FULL_IMAGE:figures/full_fig_p016_11.png]

Discussion (0). Continue with ORCID 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. DS-Det: Single-Query Paradigm and Attention Disentangled Learning for Flexible Object Detection

    cs.CV 2025-07 conditional novelty 5.0 of 10

    A flexible-query transformer detector that separates cross-attention localization from self-attention deduplication reports consistent accuracy gains over DINO across five backbones.

Reference graph

Works this paper leans on

77 extracted references · 54 canonical work pages · cited by 1 Pith paper

  1. [1]

    A full data augmentation pipeline for small object detection based on generative adversarial networks

    Brais Bosquet, Daniel Cores, Lorenzo Seidenari, V ´ıctor M Brea, Manuel Mucientes, and Alberto Del Bimbo. A full data augmentation pipeline for small object detection based on generative adversarial networks. Pattern Recognition, 133:108998, 2023

  2. [2]

    Cascade r-cnn: Delving into high quality object detection

    Zhaowei Cai and Nuno Vasconcelos. Cascade r-cnn: Delving into high quality object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 6154–6162, 2018

  3. [3]

    Visible and clear: Finding tiny objects in difference map

    Bing Cao, Haiyu Yao, Pengfei Zhu, and Qinghua Hu. Visible and clear: Finding tiny objects in difference map. In European Conference on Computer Vision, pages 1–18. Springer, 2025

  4. [4]

    Mlp-dino: Category modeling and query graphing with deep mlp for object detection

    Guiping Cao, Wenjian Huang, Xiangyuan Lan, Jianguo Zhang, Dongmei Jiang, and Yaowei Wang. Mlp-dino: Category modeling and query graphing with deep mlp for object detection. In IJCAI, 2024

  5. [5]

    Strip-mlp: Efficient token interaction for vision mlp

    Guiping Cao, Shengda Luo, Wenjian Huang, Xiangyuan Lan, Dongmei Jiang, Yaowei Wang, and Jianguo Zhang. Strip-mlp: Efficient token interaction for vision mlp. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 1494–1504, 2023

  6. [6]

    Cf-detr: Coarse-to- fine transformers for end-to-end object detection

    Xipeng Cao, Peng Yuan, Bailan Feng, and Kun Niu. Cf-detr: Coarse-to- fine transformers for end-to-end object detection. In Proceedings of the AAAI Conference on Artificial Intelligence , volume 36, pages 185–193, 2022

  7. [7]

    End-to-end object detection with transformers

    Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, and Sergey Zagoruyko. End-to-end object detection with transformers. In European conference on computer vision , pages 213–229. Springer, 2020

  8. [8]

    Crossvit: Cross-attention multi-scale vision transformer for image classification

    Chun-Fu Richard Chen, Quanfu Fan, and Rameswar Panda. Crossvit: Cross-attention multi-scale vision transformer for image classification. In Proceedings of the IEEE/CVF international conference on computer vision, pages 357–366, 2021

Show all 77 references
  1. [9]

    Diffusiondet: Diffusion model for object detection

    Shoufa Chen, Peize Sun, Yibing Song, and Ping Luo. Diffusiondet: Diffusion model for object detection. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 19830–19843, 2023

  2. [10]

    Towards large-scale small object detection: Survey and benchmarks

    Gong Cheng, Xiang Yuan, Xiwen Yao, Kebing Yan, Qinghua Zeng, Xingxing Xie, and Junwei Han. Towards large-scale small object detection: Survey and benchmarks. IEEE Transactions on Pattern Analysis and Machine Intelligence , 45(11):13467–13488, 2023

  3. [11]

    Dynamic head: Unifying object detection heads with attentions

    Xiyang Dai, Yinpeng Chen, Bin Xiao, Dongdong Chen, Mengchen Liu, Lu Yuan, and Lei Zhang. Dynamic head: Unifying object detection heads with attentions. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 7373–7382, 2021

  4. [12]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei- Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition , pages 248–255. Ieee, 2009

  5. [13]

    Pedes- trian detection: An evaluation of the state of the art

    Piotr Dollar, Christian Wojek, Bernt Schiele, and Pietro Perona. Pedes- trian detection: An evaluation of the state of the art. IEEE transactions on pattern analysis and machine intelligence , 34(4):743–761, 2011

  6. [14]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weis- senborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, et al. An image is worth 16x16 words: Transformers for image recognition at scale. In International C...

  7. [15]

    Centernet: Keypoint triplets for object detection

    Kaiwen Duan, Song Bai, Lingxi Xie, Honggang Qi, Qingming Huang, and Qi Tian. Centernet: Keypoint triplets for object detection. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6569–6578, 2019

  8. [16]

    Dssd: Deconvolutional single shot detector

    Cheng-Yang Fu, Wei Liu, Ananth Ranga, Ambrish Tyagi, and Alexan- der C Berg. Dssd: Deconvolutional single shot detector. arXiv preprint arXiv:1701.06659, 2017

  9. [17]

    Ps-rcnn: Detecting secondary human instances in a crowd via primary object suppression

    Zheng Ge, Zequn Jie, Xin Huang, Rong Xu, and Osamu Yoshie. Ps-rcnn: Detecting secondary human instances in a crowd via primary object suppression. In 2020 IEEE international conference on multimedia and expo (ICME), pages 1–6. IEEE, 2020

  10. [18]

    Save the tiny, save the all: hierarchical activation network for tiny object detection

    Guangqian Guo, Pengfei Chen, Xuehui Yu, Zhenjun Han, Qixiang Ye, and Shan Gao. Save the tiny, save the all: hierarchical activation network for tiny object detection. IEEE transactions on circuits and systems for video technology, 34:221–234, 2023

  11. [19]

    Mask r-cnn

    Kaiming He, Georgia Gkioxari, Piotr Doll ´ar, and Ross Girshick. Mask r-cnn. In Proceedings of the IEEE international conference on computer vision, pages 2961–2969, 2017

  12. [20]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 770–778, 2016

  13. [21]

    Multi- scale feature balance enhancement network for pedestrian detection

    Yuzhe He, Ning He, Ren Zhang, Kang Yan, and Haigang Yu. Multi- scale feature balance enhancement network for pedestrian detection. Multimedia Systems, 28(3):1135–1145, 2022

  14. [22]

    Gaussian error linear units (gelus)

    Dan Hendrycks and Kevin Gimpel. Gaussian error linear units (gelus). arXiv preprint arXiv:1606.08415 , 2016

  15. [23]

    Mobilenets: Efficient convolutional neural networks for mobile vision applications

    Andrew G Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, and Hartwig Adam. Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861 , 2017

  16. [24]

    Dq-detr: Detr with dynamic query for tiny object detection

    Yi-Xin Huang, Hou-I Liu, Hong-Han Shuai, and Wen-Huang Cheng. Dq-detr: Detr with dynamic query for tiny object detection. In European Conference on Computer Vision , pages 290–305. Springer, 2025

  17. [25]

    Detrs with hybrid matching

    Ding Jia, Yuhui Yuan, Haodi He, Xiaopei Wu, Haojun Yu, Weihong Lin, Lei Sun, Chao Zhang, and Han Hu. Detrs with hybrid matching. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 19702–19712, 2023

  18. [26]

    Augmentation for small object detection

    Mate Kisantal, Zbigniew Wojna, Jakub Murawski, Jacek Naruniec, and Kyunghyun Cho. Augmentation for small object detection. arXiv preprint arXiv:1902.07296, 2019

  19. [27]

    Cornernet: Detecting objects as paired keypoints

    Hei Law and Jia Deng. Cornernet: Detecting objects as paired keypoints. In Proceedings of the European conference on computer vision (ECCV) , pages 734–750, 2018

  20. [28]

    Dn-detr: Accelerate detr training by introducing query denoising

    Feng Li, Hao Zhang, Shilong Liu, Jian Guo, Lionel M Ni, and Lei Zhang. Dn-detr: Accelerate detr training by introducing query denoising. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13619–13627, 2022

  21. [29]

    Focal loss for dense object detection

    Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Doll ´ar. Focal loss for dense object detection. In Proceedings of the IEEE international conference on computer vision , pages 2980–2988, 2017

  22. [30]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceeding...

  23. [31]

    Are we ready for a new paradigm shift? a survey on visual deep mlp

    Ruiyang Liu, Yinghui Li, Linmi Tao, Dun Liang, and Hai-Tao Zheng. Are we ready for a new paradigm shift? a survey on visual deep mlp. Patterns, 3(7):100520, 2022

  24. [32]

    Dab-detr: Dynamic anchor boxes are better queries for detr

    Shilong Liu, Feng Li, Hao Zhang, Xiao Yang, Xianbiao Qi, Hang Su, Jun Zhu, and Lei Zhang. Dab-detr: Dynamic anchor boxes are better queries for detr. In International Conference on Learning Representations, 2021

  25. [33]

    Detection transformer with stable matching

    Shilong Liu, Tianhe Ren, Jiayu Chen, Zhaoyang Zeng, Hao Zhang, Feng Li, Hongyang Li, Jun Huang, Hang Su, Jun Zhu, et al. Detection transformer with stable matching. In 2023 IEEE/CVF International Conference on Computer Vision (ICCV), pages 6468–6477. IEEE, 2023

  26. [34]

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

    Shilong Liu, Zhaoyang Zeng, Tianhe Ren, Feng Li, Hao Zhang, Jie Yang, Chunyuan Li, Jianwei Yang, Hang Su, Jun Zhu, et al. Grounding dino: Marrying dino with grounded pre-training for open-set object detection. arXiv preprint arXiv:2303.05499 , 2023

  27. [35]

    Ssd: Single shot multibox detector

    Wei Liu, Dragomir Anguelov, Dumitru Erhan, Christian Szegedy, Scott Reed, Cheng-Yang Fu, and Alexander C Berg. Ssd: Single shot multibox detector. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14, p...

  28. [36]

    A survey and performance evaluation of deep learning methods for small object detection

    Yang Liu, Peng Sun, Nickolas Wergeles, and Yi Shang. A survey and performance evaluation of deep learning methods for small object detection. Expert Systems with Applications , 172:114602, 2021

  29. [37]

    Vmamba: Visual state space model

    Yue Liu, Yunjie Tian, Yuzhong Zhao, Hongtian Yu, Lingxi Xie, Yaowei Wang, Qixiang Ye, and Yunfan Liu. Vmamba: Visual state space model. arXiv preprint arXiv:2401.10166 , 2024

  30. [38]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF international conference on computer vision, pages 10012–10022, 2021

  31. [39]

    A convnet for the 2020s

    Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, and Saining Xie. A convnet for the 2020s. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11976–11986, 2022

  32. [40]

    Decoupled weight decay regulariza- tion

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regulariza- tion. In International Conference on Learning Representations , 2018

  33. [41]

    Understand- ing the effective receptive field in deep convolutional neural networks

    Wenjie Luo, Yujia Li, Raquel Urtasun, and Richard Zemel. Understand- ing the effective receptive field in deep convolutional neural networks. Advances in neural information processing systems , 29, 2016

  34. [42]

    Cascade transformer decoder based occluded pedestrian detection with dynamic deformable convolution and gaussian projection channel atten- tion mechanism

    Chunjie Ma, Li Zhuo, Jiafeng Li, Yutong Zhang, and Jing Zhang. Cascade transformer decoder based occluded pedestrian detection with dynamic deformable convolution and gaussian projection channel atten- tion mechanism. IEEE Transactions on Multimedia , 2023

  35. [43]

    Conditional detr for fast training convergence

    Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, and Jingdong Wang. Conditional detr for fast training convergence. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 3651–3660, 2021. IEEE TRANSACTIONS ON MULTIMEDIA 12

  36. [44]

    Efficient featurized image pyramid network for single shot detector

    Yanwei Pang, Tiancai Wang, Rao Muhammad Anwer, Fahad Shahbaz Khan, and Ling Shao. Efficient featurized image pyramid network for single shot detector. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 7336–7344, 2019

  37. [45]

    Detectors: Detecting objects with recursive feature pyramid and switchable atrous convolu- tion

    Siyuan Qiao, Liang-Chieh Chen, and Alan Yuille. Detectors: Detecting objects with recursive feature pyramid and switchable atrous convolu- tion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 10213–10224, 2021

  38. [46]

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

    Joseph Redmon, Santosh Divvala, Ross Girshick, and Ali Farhadi. You only look once: Unified, real-time object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 779–788, 2016

  39. [47]

    Transformers in small object detection: A benchmark and survey of state-of-the-art

    Aref Miri Rekavandi, Shima Rashidi, Farid Boussaid, Stephen Hoefs, Emre Akbas, et al. Transformers in small object detection: A benchmark and survey of state-of-the-art. arXiv preprint arXiv:2309.04902 , 2023

  40. [48]

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

    Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster r- cnn: Towards real-time object detection with region proposal networks. Advances in neural information processing systems , 28, 2015

  41. [49]

    detrex: Benchmarking detection transformers

    Tianhe Ren, Shilong Liu, Feng Li, Hao Zhang, Ailing Zeng, Jie Yang, Xingyu Liao, Ding Jia, Hongyang Li, He Cao, et al. detrex: Benchmarking detection transformers. arXiv preprint arXiv:2306.07265, 2023

  42. [50]

    Sparse detr: Efficient end-to-end object detection with learnable sparsity

    Byungseok Roh, JaeWoong Shin, Wuhyun Shin, and Saehoon Kim. Sparse detr: Efficient end-to-end object detection with learnable sparsity. In International Conference on Learning Representations , 2021

  43. [51]

    Iterdet: iterative scheme for object detection in crowded environments

    Danila Rukhovich, Konstantin Sofiiuk, Danil Galeev, Olga Barinova, and Anton Konushin. Iterdet: iterative scheme for object detection in crowded environments. In Structural, Syntactic, and Statistical Pattern Recognition: Joint IAPR International Workshops, S+ SSPR 2020, Padua...

  44. [52]

    Object detection in medical images based on hierarchical transformer and mask mechanism

    Yuntao Shou, Tao Meng, Wei Ai, Canhao Xie, Haiyan Liu, and Yina Wang. Object detection in medical images based on hierarchical transformer and mask mechanism. Computational Intelligence and Neuroscience, 2022, 2022

  45. [53]

    Sniper: Efficient multi- scale training

    Bharat Singh, Mahyar Najibi, and Larry S Davis. Sniper: Efficient multi- scale training. In Advances in Neural Information Processing Systems , volume 31. Curran Associates, Inc., 2018

  46. [54]

    Sparse r-cnn: End-to-end object detection with learnable proposals

    Peize Sun, Rufeng Zhang, Yi Jiang, Tao Kong, Chenfeng Xu, Wei Zhan, Masayoshi Tomizuka, Lei Li, Zehuan Yuan, Changhu Wang, et al. Sparse r-cnn: End-to-end object detection with learnable proposals. In Proceedings of the IEEE/CVF conference on computer vision and pattern recogn...

  47. [55]

    An image patch is a wave: Phase-aware vision mlp

    Yehui Tang, Kai Han, Jianyuan Guo, Chang Xu, Yanxi Li, Chao Xu, and Yunhe Wang. An image patch is a wave: Phase-aware vision mlp. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10935–10944, 2022

  48. [56]

    Multi- scale sampling attention graph convolutional networks for skeleton- based action recognition

    Haoyu Tian, Yipeng Zhang, Hanbo Wu, Xin Ma, and Yibin Li. Multi- scale sampling attention graph convolutional networks for skeleton- based action recognition. Neurocomputing, page 128086, 2024

  49. [57]

    Training data-efficient image transformers & distillation through attention

    Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, and Herv ´e J´egou. Training data-efficient image transformers & distillation through attention. In International conference on machine learning , pages 10347–10357. PMLR, 2021

  50. [58]

    Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors

    Chien-Yao Wang, Alexey Bochkovskiy, and Hong-Yuan Mark Liao. Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 7464–7475, 2023

  51. [59]

    A normalized gaussian wasserstein distance for tiny object detection

    Jinwang Wang, Chang Xu, Wen Yang, and Lei Yu. A normalized gaussian wasserstein distance for tiny object detection. arXiv preprint arXiv:2110.13389, 2021

  52. [60]

    Tiny object detection in aerial images

    Jinwang Wang, Wen Yang, Haowen Guo, Ruixiang Zhang, and Gui-Song Xia. Tiny object detection in aerial images. In 2020 25th international conference on pattern recognition (ICPR) , pages 3791–3798. IEEE, 2021

  53. [61]

    Convnext v2: Co-designing and scaling convnets with masked autoencoders

    Sanghyun Woo, Shoubhik Debnath, Ronghang Hu, Xinlei Chen, Zhuang Liu, In So Kweon, and Saining Xie. Convnext v2: Co-designing and scaling convnets with masked autoencoders. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 16133–16142, 2023

  54. [62]

    Rfla: Gaussian receptive field based label assignment for tiny object detection

    Chang Xu, Jinwang Wang, Wen Yang, Huai Yu, Lei Yu, and Gui-Song Xia. Rfla: Gaussian receptive field based label assignment for tiny object detection. In European conference on computer vision , pages 526–543. Springer, 2022

  55. [63]

    Querydet: Cas- caded sparse query for accelerating high-resolution small object detec- tion

    Chenhongyi Yang, Zehao Huang, and Naiyan Wang. Querydet: Cas- caded sparse query for accelerating high-resolution small object detec- tion. In Proceedings of the IEEE/CVF Conference on computer vision and pattern recognition , pages 13668–13677, 2022

  56. [64]

    Rep- points: Point set representation for object detection

    Ze Yang, Shaohui Liu, Han Hu, Liwei Wang, and Stephen Lin. Rep- points: Point set representation for object detection. In Proceedings of the IEEE/CVF international conference on computer vision , pages 9657–9666, 2019

  57. [65]

    Efficient detr: improving end-to-end object detector with dense prior

    Zhuyu Yao, Jiangbo Ai, Boxun Li, and Chi Zhang. Efficient detr: improving end-to-end object detector with dense prior. arXiv preprint arXiv:2104.01318, 2021

  58. [66]

    Multi-scale context aggregation by dilated convolutions

    Fisher Yu and Vladlen Koltun. Multi-scale context aggregation by dilated convolutions. arXiv preprint arXiv:1511.07122 , 2015

  59. [67]

    Small object detection via coarse-to-fine proposal generation and imitation learning

    Xiang Yuan, Gong Cheng, Kebing Yan, Qinghua Zeng, and Junwei Han. Small object detection via coarse-to-fine proposal generation and imitation learning. In Proceedings of the IEEE/CVF international conference on computer vision , pages 6317–6327, 2023

  60. [68]

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

    Hao Zhang, Feng Li, Shilong Liu, Lei Zhang, Hang Su, Jun Zhu, Lionel Ni, and Heung-Yeung Shum. Dino: Detr with improved denoising anchor boxes for end-to-end object detection. In The Eleventh International Conference on Learning Representations , 2022

  61. [69]

    Single-shot refinement neural network for object detection

    Shifeng Zhang, Longyin Wen, Xiao Bian, Zhen Lei, and Stan Z Li. Single-shot refinement neural network for object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4203–4212, 2018

  62. [70]

    Widerperson: A diverse dataset for dense pedestrian detection in the wild

    Shifeng Zhang, Yiliang Xie, Jun Wan, Hansheng Xia, Stan Z Li, and Guodong Guo. Widerperson: A diverse dataset for dense pedestrian detection in the wild. IEEE Transactions on Multimedia , 22(2):380– 393, 2019

  63. [71]

    Less is more: Focus attention for efficient detr

    Dehua Zheng, Wenhui Dong, Hailin Hu, Xinghao Chen, and Yunhe Wang. Less is more: Focus attention for efficient detr. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 6674–6683, 2023

  64. [72]

    Learning deep features for discriminative localization

    Bolei Zhou, Aditya Khosla, Agata Lapedriza, Aude Oliva, and Antonio Torralba. Learning deep features for discriminative localization. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 2921–2929, 2016

  65. [73]

    Detection and tracking meet drones challenge

    Pengfei Zhu, Longyin Wen, Dawei Du, Xiao Bian, Heng Fan, Qinghua Hu, and Haibin Ling. Detection and tracking meet drones challenge. IEEE Transactions on Pattern Analysis and Machine Intelligence , 44(11):7380–7399, 2021

  66. [74]

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

    Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, and Jifeng Dai. Deformable detr: Deformable transformers for end-to-end object detection. In International Conference on Learning Representations , 2020

  67. [75]

    Detrs with collaborative hy- brid assignments training

    Zhuofan Zong, Guanglu Song, and Yu Liu. Detrs with collaborative hy- brid assignments training. In Proceedings of the IEEE/CVF international conference on computer vision , pages 6748–6758, 2023

  68. [76]

    Learning data augmentation strategies for object detection

    Barret Zoph, Ekin D Cubuk, Golnaz Ghiasi, Tsung-Yi Lin, Jonathon Shlens, and Quoc V Le. Learning data augmentation strategies for object detection. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXVII 16 , pages 566–5...

  69. [77]

    Object detection in 20 years: A survey

    Zhengxia Zou, Keyan Chen, Zhenwei Shi, Yuhong Guo, and Jieping Ye. Object detection in 20 years: A survey. Proceedings of the IEEE , 111(3):257–276, 2023. IEEE TRANSACTIONS ON MULTIMEDIA 13 APPENDIX METHODOLOGY OF CLAP a) Issue of Fixed Image Size for the Deep MLP: In the main...

Pith tools

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