pith. sign in

arxiv: 2512.23221 · v1 · submitted 2025-12-29 · 💻 cs.CV · cs.AI

Holi-DETR: Holistic Fashion Item Detection Leveraging Contextual Information

Youngchae Kwon , Jinyoung Choi , Injung Kim This is my paper

Pith reviewed 2026-05-16 19:51 UTC · model grok-4.3

classification 💻 cs.CV cs.AI
keywords fashion item detectionDETRcontextual informationobject detectionco-occurrencespatial relationshipsbody keypointsholistic detection
0
0 comments X

The pith

Holi-DETR lifts fashion item detection accuracy by folding co-occurrence, spatial layout, and body-keypoint signals into the DETR framework.

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

The paper shows that standard object detectors miss relationships among clothing pieces in outfit photos, so they confuse similar items. Holi-DETR adds three explicit signals: how often items appear together, their typical relative sizes and positions, and where they sit relative to body keypoints. These signals are merged directly into the transformer layers so the model reasons about the whole outfit at once. The result is a measurable rise in average precision on fashion datasets. Readers care because reliable item detection underpins virtual try-on, search, and inventory tools that currently struggle with visual ambiguity.

Core claim

Holi-DETR detects fashion items holistically by embedding three contextual signals into the DETR architecture: co-occurrence probabilities between items, relative position and size from inter-item arrangements, and spatial links between items and human body keypoints. This integration reduces subcategory confusion that independent per-item detectors cannot resolve. Experiments report gains of 3.6 percentage points AP over vanilla DETR and 1.1 points over Co-DETR on the evaluated fashion data.

What carries the argument

The Holi-DETR architecture that fuses co-occurrence, inter-item spatial, and item-to-body-keypoint contextual features into the DETR detection pipeline.

If this is right

  • Fashion detectors become less prone to subcategory swaps when items share visual traits.
  • Outfit-level consistency improves because the model respects typical spatial and co-occurrence patterns.
  • Existing DETR and Co-DETR checkpoints can be upgraded by the same context modules without full retraining.
  • Body-keypoint alignment supplies an extra cue that pure appearance-based methods lack.

Where Pith is reading between the lines

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

  • The same three-signal pattern could transfer to scene understanding tasks where objects have stable spatial and co-occurrence rules with people.
  • If the context modules prove robust, they might lower the labeled-data requirement for new visual domains.
  • Performance on images with unusual body poses or non-standard outfits would expose the limits of the learned relationships.

Load-bearing premise

The three contextual signals can be added to DETR without creating dataset-specific biases or demanding retuning that breaks on other image types.

What would settle it

Apply the same architecture and signals to a standard object-detection benchmark outside fashion and measure whether average precision rises, stays flat, or drops.

read the original abstract

Fashion item detection is challenging due to the ambiguities introduced by the highly diverse appearances of fashion items and the similarities among item subcategories. To address this challenge, we propose a novel Holistic Detection Transformer (Holi-DETR) that detects fashion items in outfit images holistically, by leveraging contextual information. Fashion items often have meaningful relationships as they are combined to create specific styles. Unlike conventional detectors that detect each item independently, Holi-DETR detects multiple items while reducing ambiguities by leveraging three distinct types of contextual information: (1) the co-occurrence relationship between fashion items, (2) the relative position and size based on inter-item spatial arrangements, and (3) the spatial relationships between items and human body key-points. %Holi-DETR explicitly incorporates three types of contextual information: (1) the co-occurrence probability between fashion items, (2) the relative position and size based on inter-item spatial arrangements, and (3) the spatial relationships between items and human body key-points. To this end, we propose a novel architecture that integrates these three types of heterogeneous contextual information into the Detection Transformer (DETR) and its subsequent models. In experiments, the proposed methods improved the performance of the vanilla DETR and the more recently developed Co-DETR by 3.6 percent points (pp) and 1.1 pp, respectively, in terms of average precision (AP).

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

0 major / 3 minor

Summary. The manuscript proposes Holi-DETR, a Detection Transformer variant that incorporates three contextual signals—co-occurrence probabilities between fashion items, inter-item spatial relations (relative position and size), and item-to-human-keypoint spatial relations—into the decoder queries and cross-attention to enable holistic detection of fashion items in outfit images. The central claim is that this integration produces AP gains of 3.6 percentage points over vanilla DETR and 1.1 percentage points over Co-DETR under controlled comparisons using the same backbone and training protocol, supported by ablations isolating each context type.

Significance. If the gains are reproducible, the work supplies concrete evidence that domain-specific contextual priors can improve transformer detectors on structured scenes without altering the core architecture. The controlled experimental design, including ablations that isolate each of the three signals and direct comparisons against both vanilla DETR and Co-DETR, strengthens the attribution of improvements to the added context rather than training artifacts.

minor comments (3)
  1. [Abstract] The abstract reports specific AP deltas but omits the dataset name, size, and class count; moving these statistics to the abstract or the first paragraph of the introduction would improve immediate readability.
  2. [Method] A single diagram showing the precise injection points of the three context embeddings into the DETR decoder (queries and cross-attention) would clarify the architecture description and aid reproduction.
  3. [Experiments] The experimental section should report whether the AP gains are averaged over multiple random seeds or include standard deviations to confirm stability.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive assessment of Holi-DETR, the accurate summary of our contributions, and the recommendation for minor revision. The controlled comparisons and ablations isolating each contextual signal are indeed central to attributing the reported AP gains. No specific major comments were raised in the report.

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper's core contribution is an architectural modification to DETR that injects three heterogeneous context embeddings (co-occurrence probabilities, inter-item spatial relations, and item-to-keypoint relations) into the decoder queries and cross-attention layers. All reported results are empirical AP gains measured on held-out test splits against external baselines (vanilla DETR and Co-DETR) under identical backbone and training protocols, with ablations that isolate each added signal. No equation, parameter, or performance metric is defined in terms of itself or recovered by construction from the same fitted values; the derivation chain consists of standard transformer components plus explicit context encoders whose outputs are independent of the final detection scores.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The approach rests on standard DETR transformer assumptions plus the domain premise that fashion items exhibit stable co-occurrence and spatial patterns; no new free parameters or invented entities are introduced in the abstract.

axioms (2)
  • standard math DETR architecture and training assumptions hold for the augmented model
    The paper extends DETR without altering its core mathematical formulation.
  • domain assumption Fashion images contain reliable co-occurrence and spatial relationships usable as context
    Central premise stated in the abstract for why context helps.

pith-pipeline@v0.9.0 · 5550 in / 1212 out tokens · 25943 ms · 2026-05-16T19:51:05.395693+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

51 extracted references · 51 canonical work pages · 2 internal anchors

  1. [1]

    Expert Systems with Applications 116, 328 – 339 (2019)

    Yian, S., Kyungshik, S.: Hierarchical convolutional neural networks for fashion image classification. Expert systems with applications116, 328–339 (2019) https: //doi.org/10.1016/j.eswa.2018.09.022

    work page doi:10.1016/j.eswa.2018.09.022 2019

  2. [2]

    Multimedia Tools and Applications82(5), 7383–7400 (2023) https://doi.org/10.1007/s11042-022-13424-8

    Tian, Q., Chanda, S., Gray, D.: Improving apparel detection with category grouping and multi-grained branches. Multimedia Tools and Applications82(5), 7383–7400 (2023) https://doi.org/10.1007/s11042-022-13424-8

    work page doi:10.1007/s11042-022-13424-8 2023

  3. [3]

    Preprint at https://arxiv.org/abs/2111.00905 15 (2021)

    Mohammadi, S.O., Kalhor, A.: Smart fashion: a review of AI applications in the Fashion & Apparel Industry. Preprint at https://arxiv.org/abs/2111.00905 15 (2021)

    work page arXiv 2021

  4. [4]

    Sensors23(13), 6083 (2023) https://doi.org/10.3390/s23136083

    Ma, B., Xu, W.: Efficient fine tuning for fashion object detection. Sensors23(13), 6083 (2023) https://doi.org/10.3390/s23136083

    work page doi:10.3390/s23136083 2023

  5. [5]

    In: The World Wide Web Conference, pp

    Cui, Z., Li, Z., Wu, S., Zhang, X.-Y., Wang, L.: Dressing as a whole: Outfit com- patibility learning based on node-wise graph neural networks. In: The World Wide Web Conference, pp. 307–317 (2019). https://doi.org/10.1145/3308558.3313444

    work page doi:10.1145/3308558.3313444 2019

  6. [6]

    Schmon, and Chris G

    Sarkar, R., Bodla, N., Vasileva, M., Lin, Y.-L., Beniwal, A., Lu, A., Medioni, G.: Outfittransformer: Outfit representations for fashion recommendation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 2263–2267 (2022). https://doi.org/10.1109/cvprw56347.2022. 00249

    work page doi:10.1109/cvprw56347.2022 2022

  7. [7]

    In: Proceedings of the IEEE/CVF Conference on Computer 25 Vision and Pattern Recognition, pp

    Lin, Y.-L., Tran, S., Davis, L.S.: Fashion outfit complementary item retrieval. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pp. 3311–3319 (2020). https://doi.org/10.1109/cvpr42600.2020. 00337

    work page doi:10.1109/cvpr42600.2020 2020

  8. [8]

    MoCoGAN: Decomposing motion and content for video generation

    Han, X., Wu, Z., Wu, Z., Yu, R., Davis, L.S.: Viton: An image-based virtual try- on network. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 7543–7552 (2018). https://doi.org/10.1109/cvpr.2018. 00787

    work page doi:10.1109/cvpr.2018 2018

  9. [9]

    Wheat, M

    Islam, T., Miron, A., Liu, X., Li, Y.: Deep learning in virtual try-on: A comprehensive survey. IEEE Access (2024) https://doi.org/10.1109/access.2024. 3368612

    work page doi:10.1109/access.2024 2024

  10. [10]

    IEEE (pp

    Girshick, R., Donahue, J., Darrell, T., Malik, J.: Rich feature hierarchies for accurate object detection and semantic segmentation. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 580–587 (2014). https://doi.org/10.1109/CVPR.2014.81

    work page doi:10.1109/cvpr.2014.81 2014

  11. [11]

    In: 2015 IEEE International Conference on Computer Vision (ICCV), pp

    Girshick, R.: Fast r-cnn. In: 2015 IEEE International Conference on Computer Vision (ICCV), pp. 1440–1448 (2015). https://doi.org/10.1109/ICCV.2015.169

    work page doi:10.1109/iccv.2015.169 2015

  12. [12]

    IEEE transactions on pattern analy- sis and machine intelligence39(6), 1137–1149 (2016) https://doi.org/10.1109/ TPAMI.2016.2577031

    Ren, S., He, K., Girshick, R., Sun, J.: Faster r-cnn: Towards real-time object detection with region proposal networks. IEEE transactions on pattern analy- sis and machine intelligence39(6), 1137–1149 (2016) https://doi.org/10.1109/ TPAMI.2016.2577031

    work page arXiv 2016

  13. [13]

    In: Computer Vision–ECCV 2016: 14th Euro- pean Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceed- ings, Part I 14, pp

    Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C.-Y., Berg, A.C.: Ssd: Single shot multibox detector. In: Computer Vision–ECCV 2016: 14th Euro- pean Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceed- ings, Part I 14, pp. 21–37 (2016). https://doi.org/10.1007/978-3-319-46448-0 2 . Springer 16

    work page doi:10.1007/978-3-319-46448-0 2016

  14. [14]

    Communications of the ACM 65(1), 99–106 (2021) https://doi.org/ 10.1007/978-3-030-58452-8 24

    Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., Zagoruyko, S.: End-to-end object detection with transformers. In: European Conference on Com- puter Vision, pp. 213–229 (2020). https://doi.org/10.1007/978-3-030-58452-8 13 . Springer

    work page doi:10.1007/978-3-030-58452-8 2020

  15. [15]

    In: International Con- ference on Learning Representations (2021)

    Zhu, X., Su, W., Lu, L., Li, B., Wang, X., Dai, J.: Deformable detr: Deformable transformers for end-to-end object detection. In: International Con- ference on Learning Representations (2021). https://openreview.net/forum?id= gZ9hCDWe6ke

    work page 2021

  16. [16]

    Walk in the cloud: Learning curves for point clouds shape analysis, pp

    Meng, D., Chen, X., Fan, Z., Zeng, G., Li, H., Yuan, Y., Sun, L., Wang, J.: Conditional detr for fast training convergence. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 3651–3660 (2021). https://doi. org/10.1109/ICCV48922.2021.00363

    work page doi:10.1109/iccv48922.2021.00363 2021

  17. [17]

    Advances in Neural Information Pro- cessing Systems35, 34899–34911 (2022) https://doi.org/10.48550/arXiv.2205

    Tian, Z., Chu, X., Wang, X., Wei, X., Shen, C.: Fully convolutional one-stage 3d object detection on lidar range images. Advances in Neural Information Pro- cessing Systems35, 34899–34911 (2022) https://doi.org/10.48550/arXiv.2205. 13764

    work page doi:10.48550/arxiv.2205 2022

  18. [18]

    In: Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pp

    Li, F., Zhang, H., Liu, S., Guo, J., Ni, L.M., Zhang, L.: Dn-detr: Accelerate detr training by introducing query denoising. In: Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pp. 13619–13627 (2022). https://doi.org/10.1109/tpami.2023.3335410

    work page doi:10.1109/tpami.2023.3335410 2022

  19. [19]

    Dickerson

    Liu, S., Li, F., Zhang, H., Yang, X., Qi, X., Su, H., Zhu, J., Zhang, L.: DAB- DETR: Dynamic anchor boxes are better queries for DETR. In: International Conference on Learning Representations (2022). https://doi.org/10.48550/arXiv. 2201.12329 . https://openreview.net/forum?id=oMI9PjOb9Jl

    work page internal anchor Pith review doi:10.48550/arxiv 2022

  20. [20]

    DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection

    Zhang, H., Li, F., Liu, S., Zhang, L., Su, H., Zhu, J., Ni, L., Shum, H.-Y.: Dino: Detr with improved denoising anchor boxes for end-to-end object detection. In: The Eleventh International Conference on Learning Representations (2022). https://doi.org/10.48550/arXiv.2203.03605

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2203.03605 2022

  21. [21]

    Preprint at https://arxiv.org/abs/2306.04670 (2023)

    Shehzadi, T., Hashmi, K.A., Stricker, D., Afzal, M.Z.: 2d object detection with transformers: a review. Preprint at https://arxiv.org/abs/2306.04670 (2023). https://doi.org/10.48550/arXiv.2306.04670

    work page doi:10.48550/arxiv.2306.04670 2023

  22. [22]

    In: European Conference on Computer Vision, pp

    Hou, X., Liu, M., Zhang, S., Wei, P., Chen, B., Lan, X.: Relation detr: Exploring explicit position relation prior for object detection. In: European Conference on Computer Vision, pp. 89–105 (2025). https://doi.org/10.1007/ 978-3-031-72973-7 6 . Springer

    work page 2025

  23. [23]

    CCCV 2015 Comput

    Lao, B., Jagadeesh, K.: Convolutional neural networks for fashion classification and object detection. CCCV 2015 Comput. Vis546, 120–129 (2015) 17

    work page 2015

  24. [24]

    In: 2018 IEEE 4th International Conference on Computer and Communications (ICCC), pp

    Feng, Z., Luo, X., Yang, T., Kita, K.: An object detection system based on yolov2 in fashion apparel. In: 2018 IEEE 4th International Conference on Computer and Communications (ICCC), pp. 1532–1536 (2018). https://doi.org/10.1109/ compcomm.2018.8780944 . IEEE

    work page arXiv 2018

  25. [25]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp

    Duan, K., Bai, S., Xie, L., Qi, H., Huang, Q., Tian, Q.: Centernet: Keypoint triplets for object detection. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 6569–6578 (2019). https://doi.org/10.1109/ ICCV.2019.00667

    work page arXiv 2019

  26. [26]

    Multi-Layout Unstructured Invoice Documents Dataset: A Dataset for Template-Free Invoice Processing and Its Evaluation Using AI Ap- proaches

    Kim, H.J., Lee, D.H., Niaz, A., Kim, C.Y., Memon, A.A., Choi, K.N.: Multiple- clothing detection and fashion landmark estimation using a single-stage detec- tor. IEEE Access9, 11694–11704 (2021) https://doi.org/10.1109/access.2021. 3051424

    work page doi:10.1109/access.2021 2021

  27. [27]

    Applied Sciences11(9), 3782 (2021) https://doi.org/10.3390/ app11093782

    Lee, C.-H., Lin, C.-W.: A two-phase fashion apparel detection method based on yolov4. Applied Sciences11(9), 3782 (2021) https://doi.org/10.3390/ app11093782

    work page 2021

  28. [28]

    Efficient attention: Attention with linear complexities

    Sidnev, A., Krapivin, A., Trushkov, A., Krasikova, E., Kazakov, M., Viryasov, M.: Deepmark++: Real-time clothing detection at the edge. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pp. 2980– 2988 (2021). https://doi.org/10.1109/wacv48630.2021.00302

    work page doi:10.1109/wacv48630.2021.00302 2021

  29. [29]

    In: Journal of Physics: Conference Series, vol

    Alamsyah, A., Saputra, M.A.A., Masrury, R.A.: Object detection using con- volutional neural network to identify popular fashion product. In: Journal of Physics: Conference Series, vol. 1192, p. 012040 (2019). https://doi.org/10.1088/ 1742-6596/1192/1/012040 . IOP Publishing

    work page 2019

  30. [30]

    The Visual Computer, 1–13 (2024) https://doi.org/10.1007/s00371-024-03337-y

    Li, Y., Zhang, W., Wu, M., Zhang, D., Wang, Z., You, C.: Multi-keypoints matching network for clothing detection. The Visual Computer, 1–13 (2024) https://doi.org/10.1007/s00371-024-03337-y

    work page doi:10.1007/s00371-024-03337-y 2024

  31. [31]

    In: 2016 IEEE Winter Conference on Applications of Computer Vision (WACV), pp

    Hara, K., Jagadeesh, V., Piramuthu, R.: Fashion apparel detection: the role of deep convolutional neural network and pose-dependent priors. In: 2016 IEEE Winter Conference on Applications of Computer Vision (WACV), pp. 1–9 (2016). https://doi.org/10.1109/WACV.2016.7477611 . IEEE

    work page doi:10.1109/wacv.2016.7477611 2016

  32. [32]

    In: 2007 IEEE 11th International Conference on Computer Vision, pp

    Rabinovich, A., Vedaldi, A., Galleguillos, C., Wiewiora, E., Belongie, S.: Objects in context. In: 2007 IEEE 11th International Conference on Computer Vision, pp. 1–8 (2007). https://doi.org/10.1109/ICCV.2007.4408986 . IEEE

    work page doi:10.1109/iccv.2007.4408986 2007

  33. [34]

    In: 2023 IEEE/CVF International Conference on Computer Vision (ICCV)

    Zong, Z., Song, G., Liu, Y.: Detrs with collaborative hybrid assignments training. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 6748–6758 (2023). https://doi.org/10.1109/iccv51070.2023.00621

    work page doi:10.1109/iccv51070.2023.00621 2023

  34. [35]

    2016, in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 779–788, doi: 10.1109/CVPR.2016.91

    Redmon, J.: You only look once: Unified, real-time object detection. In: Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2016). https://doi.org/10.1109/CVPR.2016.91

    work page doi:10.1109/cvpr.2016.91 2016

  35. [36]

    In: International Conference on Learning Representations (2021)

    Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., Houlsby, N.: An image is worth 16x16 words: Transformers for image recognition at scale. In: International Conference on Learning Representations (2021). https: //openreview.net/forum?id=YicbFdNTTy

    work page 2021

  36. [37]

    Walk in the cloud: Learning curves for point clouds shape analysis, pp

    Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., Guo, B.: Swin trans- former: Hierarchical vision transformer using shifted windows. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 10012–10022 (2021). https://doi.org/10.1109/ICCV48922.2021.00986

    work page doi:10.1109/iccv48922.2021.00986 2021

  37. [38]

    In: Computer Vision–ECCV 2012: 12th European Conference on Computer Vision, Florence, Italy, October 7-13, 2012, Proceedings, Part VI 12, pp

    Cinbis, R.G., Sclaroff, S.: Contextual object detection using set-based classifica- tion. In: Computer Vision–ECCV 2012: 12th European Conference on Computer Vision, Florence, Italy, October 7-13, 2012, Proceedings, Part VI 12, pp. 43–57 (2012). https://doi.org/10.1007/978-3-642-33783-3 4 . Springer

    work page doi:10.1007/978-3-642-33783-3 2012

  38. [39]

    IEEE Transactions on Cognitive and Developmental Systems14(4), 1320–1330 (2020) https://doi.org/10.1109/TCDS.2020.3008213

    Alamri, F., Pugeault, N.: Improving object detection performance using scene contextual constraints. IEEE Transactions on Cognitive and Developmental Systems14(4), 1320–1330 (2020) https://doi.org/10.1109/TCDS.2020.3008213

    work page doi:10.1109/tcds.2020.3008213 2020

  39. [40]

    Computer vision and image understanding114(6), 712–722 (2010) https: //doi.org/10.1016/j.cviu.2010.02.004

    Galleguillos, C., Belongie, S.: Context based object categorization: A critical sur- vey. Computer vision and image understanding114(6), 712–722 (2010) https: //doi.org/10.1016/j.cviu.2010.02.004

    work page doi:10.1016/j.cviu.2010.02.004 2010

  40. [41]

    2016.280

    Galleguillos, C., Rabinovich, A., Belongie, S.: Object categorization using co- occurrence, location and appearance. In: 2008 IEEE Conference on Computer Vision and Pattern Recognition, pp. 1–8 (2008). https://doi.org/10.1109/CVPR. 2008.4587799 . IEEE

    work page doi:10.1109/cvpr 2008

  41. [42]

    In: WSCG 2016 - 24th Conference on Computer Graphics, Visu- alization and Computer Vision (2016)

    Zolghadr, E., Furht, B.: Scene understanding using context-based conditional random field. In: WSCG 2016 - 24th Conference on Computer Graphics, Visu- alization and Computer Vision (2016). https://doi.org/https://doi.org/wscg.zcu. cz/WSCG2016/!! CSRN-2601.pdf

    work page 2016

  42. [43]

    In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp

    Hu, H., Gu, J., Zhang, Z., Dai, J., Wei, Y.: Relation networks for object detec- tion. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 3588–3597 (2018). https://doi.org/10.1109/cvpr.2018.00378

    work page doi:10.1109/cvpr.2018.00378 2018

  43. [44]

    In: Computer Vision–ACCV 2018: 14th Asian Conference on Com- puter Vision, Perth, Australia, December 2–6, 2018, Revised Selected Papers, Part II 14, pp

    Barnea, E., Ben-Shahar, O.: Contextual object detection with a few relevant 19 neighbors. In: Computer Vision–ACCV 2018: 14th Asian Conference on Com- puter Vision, Perth, Australia, December 2–6, 2018, Revised Selected Papers, Part II 14, pp. 480–495 (2019). https://doi.org/10.1007/978-3-030-20890-5 31 . Springer

    work page doi:10.1007/978-3-030-20890-5 2018

  44. [45]

    In: 2019 Joint IEEE 9th International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob), pp

    Alamri, F., Pugeault, N.: Contextual relabelling of detected objects. In: 2019 Joint IEEE 9th International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob), pp. 313–319 (2019). https://doi.org/10.1109/devlrn. 2019.8850686 . IEEE

    work page doi:10.1109/devlrn 2019

  45. [46]

    In: Proceedings of the IEEE/CVF Conference on Computer 25 Vision and Pattern Recognition, pp

    Pato, L.V., Negrinho, R., Aguiar, P.M.: Seeing without looking: Contextual rescoring of object detections for ap maximization. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 14610–14618 (2020). https://doi.org/10.1109/cvpr42600.2020.01462

    work page doi:10.1109/cvpr42600.2020.01462 2020

  46. [47]

    In: Proceedings of the Thirty- Second International Joint Conference on Artificial Intelligence, pp

    Hao, X., Huang, D., Lin, J., Lin, C.-Y.: Relation-enhanced detr for component detection in graphic design reverse engineering. In: Proceedings of the Thirty- Second International Joint Conference on Artificial Intelligence, pp. 4785–4793 (2023). https://doi.org/10.24963/ijcai.2023/532

    work page doi:10.24963/ijcai.2023/532 2023

  47. [48]

    Advances in Neural Information Pro- cessing Systems35, 38571–38584 (2022) https://doi.org/10.1109/iccc56324.2022

    Xu, Y., Zhang, J., Zhang, Q., Tao, D.: Vitpose: Simple vision transformer baselines for human pose estimation. Advances in Neural Information Pro- cessing Systems35, 38571–38584 (2022) https://doi.org/10.1109/iccc56324.2022. 10065997

    work page doi:10.1109/iccc56324.2022 2022

  48. [49]

    Scale-Free Networks: Complex Webs in Nature and Technology

    Church, K.W., Hanks, P.: Word association norms, mutual information, and lexicography. Computational Linguistics16(1), 22–29 (1990) https://doi.org/10. 1093/oso/9780199292332.003.0019

    work page arXiv 1990

  49. [50]

    https://github.com/facebookresearch/detr

    Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., Zagoruyko, S.: DETR (DEtection TRansformer). https://github.com/facebookresearch/detr

    work page

  50. [51]

    https://github.com/ Sense-X/Co-DETR

    Sense-X: Co-DETR: Cooperative Detection Transformer. https://github.com/ Sense-X/Co-DETR

    work page

  51. [52]

    Deep residual learning for image recognition,

    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recogni- tion. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778 (2016). https://doi.org/10.1109/CVPR.2016.90 20

    work page doi:10.1109/cvpr.2016.90 2016