pith. machine review for the scientific record. sign in

arxiv: 2604.20281 · v2 · submitted 2026-04-22 · 💻 cs.CV

Fourier Series Coder: A Novel Perspective on Angle Boundary Discontinuity Problem for Oriented Object Detection

Minghong Wei , Pu Cao , Zhihao Chen , Zhiyuan Zang , Lu Yang , Qing Song This is my paper

Pith reviewed 2026-05-10 00:25 UTC · model grok-4.3

classification 💻 cs.CV
keywords oriented object detectionangle boundary discontinuityFourier series coderangle encodingcyclic ambiguityphase unwrappingmanifold constraint
0
0 comments X

The pith

Mapping angles to a minimal orthogonal Fourier basis with a manifold constraint eliminates cyclic errors in oriented object detection.

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

The paper aims to solve the angle boundary discontinuity and cyclic ambiguity problems that cause fluctuations in oriented object detection for applications like autonomous driving and remote sensing. It identifies that state-of-the-art continuous angle coders still produce substantial cyclic errors, which it traces to noise amplification in non-orthogonal decoding. The proposed Fourier Series Coder resolves this by mapping angles onto a minimal orthogonal Fourier basis while enforcing a geometric manifold constraint. This approach prevents feature modulus collapse and delivers robust phase unwrapping without relying on heuristic truncations. The result is strict boundary continuity, better noise immunity, and measurable gains in high-precision detection across large datasets.

Core claim

By rigorously mapping angles onto a minimal orthogonal Fourier basis and explicitly enforcing a geometric manifold constraint, the Fourier Series Coder prevents feature modulus collapse. This structurally stabilized representation ensures highly robust phase unwrapping, intrinsically eliminating the need for heuristic truncations while achieving strict boundary continuity and superior noise immunity.

What carries the argument

The Fourier Series Coder, a lightweight plug-and-play module that maps angles to a minimal orthogonal Fourier basis while enforcing a geometric manifold constraint to stabilize the encoding-decoding process.

If this is right

  • Strict boundary continuity holds across all periodic angle boundaries without manual fixes.
  • No heuristic truncations are required during phase unwrapping.
  • Noise immunity improves especially for square-like objects that previously amplified errors.
  • High-precision detection metrics rise substantially on large-scale datasets.
  • The coder integrates as a drop-in replacement in existing oriented detectors.

Where Pith is reading between the lines

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

  • The same orthogonal-basis stabilization could apply to other periodic regression tasks such as pose estimation or rotation-invariant feature learning.
  • Detectors handling arbitrary aspect ratios might adopt similar manifold constraints to reduce sensitivity to object shape.
  • Cross-dataset tests on more extreme aspect ratios could expose whether the manifold constraint remains sufficient outside the evaluated domains.

Load-bearing premise

That non-orthogonal decoding noise amplification is the root cause of cyclic errors in prior coders, and that an orthogonal Fourier mapping plus manifold constraint will eliminate these errors for all object shapes and datasets without introducing new instabilities.

What would settle it

Persistent angle fluctuations near periodic boundaries for square-like objects on any of the three tested large-scale datasets would show that the orthogonal mapping and manifold constraint have not achieved strict boundary continuity.

Figures

Figures reproduced from arXiv: 2604.20281 by Lu Yang, Minghong Wei, Pu Cao, Qing Song, Zhihao Chen, Zhiyuan Zang.

Figure 1
Figure 1. Figure 1: Examples of discontinuities caused by rotational symmetry. (a) Angle [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Comprehensive comparison of decoding stability under predictive noise. (a) and (b) present the discrete angle predictions for a rectangular and a [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The optimization paradigm of the proposed Fourier Series Coder. (a) The OBB coder encodes the angle with other parameters together and uses [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The angle mapping functions for rectangular and square objects. Our [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Compare the angle prediction errors of three angle encoders: FSC [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Comparison of component fluctuations between PSC and FSC (n=1). [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Visual analysis of the manifold constraint. (a) Raw PSCD components [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Visual comparison of three angle coders on the DOTA-v1.0 dataset. [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Illustration of the angle results with different angle coder settings. Green dots indicate correct detection within the allowable range, while purple dots [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Visual comparison of KLD [18] and KLD+ (ours) for detecting square-like objects.. TABLE VI COMPARISON WITH GAUSSIAN-BASED METHODS W/ AND W/O FSC IN TERMS OF IN AP (%) AND PARAMETERS (M) ON THE DIOR-R DATASET. Method AP50 AP75 Params GWD 54.80 34.00 36.52 GWD+ 58.90 (+4.10) 39.00 (+5.00) 36.60 (+0.08) KLD 55.60 34.40 36.52 KLD+ 59.80 (+4.20) 37.70 (+3.30) 36.60 (+0.08) spatial parameters (x, y, w, h), whil… view at source ↗
read the original abstract

With the rapid advancement of intelligent driving and remote sensing, oriented object detection has gained widespread attention. However, achieving high-precision performance is fundamentally constrained by the Angle Boundary Discontinuity (ABD) and Cyclic Ambiguity (CA) problems, which typically cause significant angle fluctuations near periodic boundaries. Although recent studies propose continuous angle coders to alleviate these issues, our theoretical and empirical analyses reveal that state-of-the-art methods still suffer from substantial cyclic errors. We attribute this instability to the structural noise amplification within their non-orthogonal decoding mechanisms. This mathematical vulnerability significantly exacerbates angular deviations, particularly for square-like objects. To resolve this fundamentally, we propose the Fourier Series Coder (FSC), a lightweight plug-and-play component that establishes a continuous, reversible, and mathematically robust angle encoding-decoding paradigm. By rigorously mapping angles onto a minimal orthogonal Fourier basis and explicitly enforcing a geometric manifold constraint, FSC effectively prevents feature modulus collapse. This structurally stabilized representation ensures highly robust phase unwrapping, intrinsically eliminating the need for heuristic truncations while achieving strict boundary continuity and superior noise immunity. Extensive experiments across three large-scale datasets demonstrate that FSC achieves highly competitive overall performance, yielding substantial improvements in high-precision detection. The code will be available at https://github.com/weiminghong/FSC.

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 / 2 minor

Summary. The manuscript proposes the Fourier Series Coder (FSC) as a solution to the Angle Boundary Discontinuity (ABD) and Cyclic Ambiguity (CA) problems in oriented object detection. It argues that existing continuous angle coders exhibit cyclic errors due to noise amplification in non-orthogonal decoding, particularly for square-like objects. FSC maps angles to a minimal orthogonal Fourier basis while enforcing a geometric manifold constraint to prevent feature modulus collapse, enabling robust phase unwrapping without heuristic truncations. The approach is claimed to achieve strict boundary continuity and superior noise immunity. Experiments on three large-scale datasets demonstrate competitive performance with substantial gains in high-precision detection.

Significance. If the claims hold, this work provides a mathematically principled and lightweight method to address a core challenge in oriented object detection. The emphasis on orthogonal encoding and manifold constraints offers a structural fix rather than heuristic patches, which could improve robustness in applications like autonomous driving and remote sensing. The plug-and-play nature and promise of code release are positive for reproducibility.

minor comments (2)
  1. [Abstract] Abstract: The abstract refers to 'theoretical and empirical analyses' revealing noise amplification but does not preview any key equations or specific error metrics; including a brief reference to the core mathematical insight (e.g., the orthogonal basis mapping) would strengthen the summary for readers.
  2. [Experiments] Experiments section: The manuscript mentions experiments across three datasets yielding 'substantial improvements' but the provided abstract lacks quantitative details, ablation studies, or error-bar evidence; ensure the full paper includes these to support the high-precision detection claims.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive and accurate summary of our work, as well as the recommendation for minor revision. The referee correctly identifies the core motivation—addressing angle boundary discontinuity and cyclic ambiguity via an orthogonal Fourier basis with manifold constraints—and notes the plug-and-play nature and reproducibility benefits. Since the report lists no specific major comments, we have no point-by-point rebuttals to provide. We will incorporate any minor suggestions in the revised version and release the code as stated.

Circularity Check

0 steps flagged

No significant circularity; derivation uses standard Fourier properties plus novel constraint

full rationale

The paper's central construction maps periodic angles to a minimal orthogonal Fourier basis and adds an explicit geometric manifold constraint to prevent modulus collapse and ensure stable phase unwrapping. This rests on well-established Fourier-series orthogonality (external to the paper) and a newly introduced constraint rather than any fitted parameter, self-cited uniqueness theorem, or reduction of a prediction to its own inputs. No equation or claim in the provided abstract or reader summary reduces by construction to prior author results; the attribution of prior cyclic errors to non-orthogonal decoding is a hypothesis that the new encoding directly addresses without circularity. The overall argument is self-contained against external mathematical benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

The central claim depends on standard properties of Fourier series for periodic functions and on the unproven assertion that orthogonality plus a geometric manifold constraint will prevent modulus collapse and guarantee robust unwrapping.

axioms (2)
  • standard math Fourier series provide a continuous, orthogonal representation of periodic angle functions
    Invoked to map angles onto a minimal basis that avoids boundary jumps.
  • domain assumption Enforcing a geometric manifold constraint on the encoded features prevents modulus collapse
    This is the key stability mechanism asserted but not derived from first principles in the abstract.
invented entities (1)
  • Fourier Series Coder (FSC) no independent evidence
    purpose: Lightweight plug-and-play angle encoding-decoding module
    New component introduced to replace prior non-orthogonal coders.

pith-pipeline@v0.9.0 · 5543 in / 1366 out tokens · 39529 ms · 2026-05-10T00:25:55.639837+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

56 extracted references · 3 canonical work pages

  1. [1]

    Learning roi trans- former for oriented object detection in aerial images,

    J. Ding, N. Xue, Y . Long, G.-S. Xia, and Q. Lu, “Learning roi trans- former for oriented object detection in aerial images,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 2849–2858. 1

    2019

  2. [2]

    Align deep features for oriented object detection,

    J. Han, J. Ding, J. Li, and G.-S. Xia, “Align deep features for oriented object detection,”IEEE transactions on geoscience and remote sensing, vol. 60, pp. 1–11, 2021. 1, 6, 10

    2021

  3. [3]

    R3det: Refined single-stage detector with feature refinement for rotating object,

    X. Yang, J. Yan, Z. Feng, and T. He, “R3det: Refined single-stage detector with feature refinement for rotating object,” inProceedings of the AAAI conference on artificial intelligence, vol. 35, no. 4, 2021, pp. 3163–3171. 1, 10

    2021

  4. [4]

    Cfc-net: A critical feature capturing network for arbitrary-oriented object detection in remote- sensing images,

    Q. Ming, L. Miao, Z. Zhou, and Y . Dong, “Cfc-net: A critical feature capturing network for arbitrary-oriented object detection in remote- sensing images,”IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1–14, 2021. 1

    2021

  5. [5]

    Wholly-wood: Wholly leveraging diversified-quality labels for weakly-supervised oriented ob- ject detection,

    Y . Yu, X. Yang, Y . Li, Z. Han, F. Da, and J. Yan, “Wholly-wood: Wholly leveraging diversified-quality labels for weakly-supervised oriented ob- ject detection,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025. 1

    2025

  6. [6]

    S3inet: Semantic-information space sharing in- teraction network for arbitrary shape text detection,

    R. Wang, H. Chen, Y . Zhu, J. Xu, X. Cao, Z. Zhu, S. Qian, C. Gao, L. Liu, and N. Sang, “S3inet: Semantic-information space sharing in- teraction network for arbitrary shape text detection,”IEEE Transactions on Neural Networks and Learning Systems, 2025. 1

    2025

  7. [7]

    Rotation-sensitive regression for oriented scene text detection,

    M. Liao, Z. Zhu, B. Shi, G.-s. Xia, and X. Bai, “Rotation-sensitive regression for oriented scene text detection,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 5909–

    2018

  8. [8]

    Arbitrary-oriented scene text detection via rotation proposals,

    J. Ma, W. Shao, H. Ye, L. Wang, H. Wang, Y . Zheng, and X. Xue, “Arbitrary-oriented scene text detection via rotation proposals,”IEEE transactions on multimedia, vol. 20, no. 11, pp. 3111–3122, 2018. 1

    2018

  9. [9]

    Efficient defect detection of rotating goods under the background of intelligent retail,

    Z. Hu, X. Zeng, K. Xie, C. Wen, J. He, and W. Zhang, “Efficient defect detection of rotating goods under the background of intelligent retail,” Sensors, vol. 24, no. 2, 2024. 1

    2024

  10. [10]

    Real-world multiobject, multigrasp detection,

    F.-J. Chu, R. Xu, and P. A. Vela, “Real-world multiobject, multigrasp detection,”IEEE Robotics and Automation Letters, vol. 3, no. 4, pp. 3355–3362, 2018. 1

    2018

  11. [11]

    Antipodal robotic grasping using generative residual convolutional neural network,

    S. Kumra, S. Joshi, and F. Sahin, “Antipodal robotic grasping using generative residual convolutional neural network,” in2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2020, pp. 9626–9633. 1

    2020

  12. [12]

    Arbitrary-oriented object detection with circular smooth label,

    X. Yang and J. Yan, “Arbitrary-oriented object detection with circular smooth label,” inComputer Vision–ECCV 2020: 16th European Con- ference, Glasgow, UK, August 23–28, 2020, Proceedings, Part VIII 16. Springer, 2020, pp. 677–694. 1, 3, 4, 10

    2020

  13. [13]

    Learning modulated loss for rotated object detection,

    W. Qian, X. Yang, S. Peng, J. Yan, and Y . Guo, “Learning modulated loss for rotated object detection,” inProceedings of the AAAI conference on artificial intelligence, vol. 35, no. 3, 2021, pp. 2458–2466. 1, 2

    2021

  14. [14]

    Scrdet: Towards more robust detection for small, cluttered and rotated objects,

    X. Yang, J. Yang, J. Yan, Y . Zhang, T. Zhang, Z. Guo, X. Sun, and K. Fu, “Scrdet: Towards more robust detection for small, cluttered and rotated objects,” inProceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 8232–8241. 1, 2

    2019

  15. [15]

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

    G. Nie and H. Huang, “Multi-oriented object detection in aerial images with double horizontal rectangles,”IEEE Transactions on Pattern Anal- ysis and Machine Intelligence, vol. 45, no. 4, pp. 4932–4944, 2022. 1, 3

    2022

  16. [16]

    Phase-shifting coder: Predicting accurate orientation in oriented object detection,

    Y . Yu and F. Da, “Phase-shifting coder: Predicting accurate orientation in oriented object detection,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 13 354–13 363. 1, 2, 3, 4, 5, 10 12

    2023

  17. [17]

    Rethinking rotated object detection with gaussian wasserstein distance loss,

    X. Yang, J. Yan, Q. Ming, W. Wang, X. Zhang, and Q. Tian, “Rethinking rotated object detection with gaussian wasserstein distance loss,” in International conference on machine learning. PMLR, 2021, pp. 11 830–11 841. 1, 3, 10

    2021

  18. [18]

    Learning high-precision bounding box for rotated object detection via kullback-leibler divergence,

    X. Yang, X. Yang, J. Yang, Q. Ming, W. Wang, Q. Tian, and J. Yan, “Learning high-precision bounding box for rotated object detection via kullback-leibler divergence,”Advances in Neural Information Process- ing Systems, vol. 34, pp. 18 381–18 394, 2021. 1, 3, 4, 10

    2021

  19. [19]

    The kfiou loss for rotated object detection,

    X. Yang, Y . Zhou, G. Zhang, J. Yang, W. Wang, J. Yan, X. Zhang, and Q. Tian, “The kfiou loss for rotated object detection,”arXiv preprint arXiv:2201.12558, 2022. 1, 3, 10

    work page arXiv 2022

  20. [20]

    Gaucho: Gaussian distributions with cholesky decomposition for oriented object detection,

    J. H. L. Marques, J. Murrugarra-Llerena, and C. R. Jung, “Gaucho: Gaussian distributions with cholesky decomposition for oriented object detection,” inProceedings of the Computer Vision and Pattern Recog- nition Conference, 2025, pp. 3593–3602. 1, 10

    2025

  21. [21]

    Gsdet: Gaussian splatting for oriented object detection,

    Z. Ding, J. Zhao, Y . Zhou, W.-l. Du, H. Zhu, and R. Yao, “Gsdet: Gaussian splatting for oriented object detection,” 2025. 1, 10

    2025

  22. [22]

    A novel object detector based on high-quality rotation proposal generation and adaptive angle optimization,

    Y . Qiao, L. Miao, Z. Zhou, and Q. Ming, “A novel object detector based on high-quality rotation proposal generation and adaptive angle optimization,”IEEE Transactions on Geoscience and Remote Sensing, vol. 61, pp. 1–15, 2023. 1, 3

    2023

  23. [23]

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

    Y . Xu, M. Fu, Q. Wang, Y . Wang, K. Chen, G.-S. Xia, and X. Bai, “Gliding vertex on the horizontal bounding box for multi-oriented object detection,”IEEE transactions on pattern analysis and machine intelligence, vol. 43, no. 4, pp. 1452–1459, 2020. 1, 3, 10

    2020

  24. [24]

    Theoretically achieving continuous representation of oriented bounding boxes,

    Z. Xiao, G. Yang, X. Yang, T. Mu, J. Yan, and S. Hu, “Theoretically achieving continuous representation of oriented bounding boxes,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 16 912–16 922. 1, 3

    2024

  25. [25]

    Adaptive period embedding for representing oriented objects in aerial images,

    Y . Zhu, J. Du, and X. Wu, “Adaptive period embedding for representing oriented objects in aerial images,”IEEE Transactions on Geoscience and Remote Sensing, vol. 58, no. 10, pp. 7247–7257, 2020. 1

    2020

  26. [26]

    Arbitrary- oriented object detection in remote sensing images based on polar coordinates,

    L. Zhou, H. Wei, H. Li, W. Zhao, Y . Zhang, and Y . Zhang, “Arbitrary- oriented object detection in remote sensing images based on polar coordinates,”IEEE Access, vol. 8, pp. 223 373–223 384, 2020. 1

    2020

  27. [27]

    Gaussian label distribution learning for spherical image object detection,

    H. Xu, X. Liu, Q. Zhao, Y . Ma, C. Yan, and F. Dai, “Gaussian label distribution learning for spherical image object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 1033–1042. 1

    2023

  28. [28]

    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,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 15 819–15 829. 1, 3, 10

    2021

  29. [29]

    Ars-detr: Aspect ratio sensitive oriented object detection with transformer,

    Y . Zeng, L. Q. Yang, Xue, Y . Chen, and J. Yan, “Ars-detr: Aspect ratio sensitive oriented object detection with transformer,”arXiv preprint arXiv:2303.04989, 2023. 1, 3

    work page arXiv 2023

  30. [30]

    Multigrained angle representation for remote-sensing object detection,

    H. Wang, Z. Huang, Z. Chen, Y . Song, and W. Li, “Multigrained angle representation for remote-sensing object detection,”IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1–13, 2022. 1, 3

    2022

  31. [31]

    Piou loss: Towards accurate oriented object detection in complex environments,

    Z. Chen, K. Chen, W. Lin, J. See, H. Yu, Y . Ke, and C. Yang, “Piou loss: Towards accurate oriented object detection in complex environments,” inEuropean Conference on Computer Vision. Springer, 2020, pp. 195–

    2020

  32. [32]

    Probabilistic intersection-over-union for training and evaluation of oriented object detectors,

    J. Murrugarra-Llerena, L. N. Kirsten, L. F. Zeni, and C. R. Jung, “Probabilistic intersection-over-union for training and evaluation of oriented object detectors,”IEEE Transactions on Image Processing, vol. 33, pp. 671–681, 2024. 3

    2024

  33. [33]

    Gaussian focal loss: Learning distribution polarized angle prediction for rotated object detection in aerial images,

    J. Wang, F. Li, and H. Bi, “Gaussian focal loss: Learning distribution polarized angle prediction for rotated object detection in aerial images,” IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1– 13, 2022. 3

    2022

  34. [34]

    Rethinking boundary discontinuity problem for oriented object detection,

    H. Xu, X. Liu, H. Xu, Y . Ma, Z. Zhu, C. Yan, and F. Dai, “Rethinking boundary discontinuity problem for oriented object detection,” inPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 17 406–17 415. 3, 10

    2024

  35. [35]

    Trigonometric-coded refined detector for high precision oriented object detection,

    R. Zhang, Y . Wang, S. Shen, W. Zhao, Z. Zeng, N. Li, and D. Li, “Trigonometric-coded refined detector for high precision oriented object detection,”IEEE Geoscience and Remote Sensing Letters, vol. 20, pp. 1–5, 2023. 3

    2023

  36. [36]

    Coupled dual-frequency phase-shifting coder for precise rotated angle representa- tion in oriented object detection,

    F. Liu, R. Chen, J. Zhang, S. Ding, H. Liu, S. Ma, and K. Xing, “Coupled dual-frequency phase-shifting coder for precise rotated angle representa- tion in oriented object detection,”IEEE Geoscience and Remote Sensing Letters, vol. 21, pp. 1–5, 2024. 3

    2024

  37. [37]

    Direction prediction redefinition: transfer angle to scale in oriented object detection,

    B. Song, J. Li, J. Wu, J. Chang, and J. Wan, “Direction prediction redefinition: transfer angle to scale in oriented object detection,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 34, no. 12, pp. 12 894–12 906, 2024. 3

    2024

  38. [38]

    Hierarchical mask prompting and robust integrated regression for oriented object de- tection,

    Y . Yao, G. Cheng, C. Lang, X. Yuan, X. Xie, and J. Han, “Hierarchical mask prompting and robust integrated regression for oriented object de- tection,”IEEE transactions on circuits and systems for video technology, vol. 34, no. 12, pp. 13 071–13 084, 2024. 3

    2024

  39. [39]

    Fourier angle alignment for oriented object detection in remote sensing,

    C. Gu, L. Chen, L. Gu, and Y . Fu, “Fourier angle alignment for oriented object detection in remote sensing,”arXiv preprint arXiv:2602.23790,

    work page arXiv

  40. [40]

    Efficient oriented object detection via wavelet-based energy label reassignment and dual prediction strategy,

    B. Song, J. Li, J. Wu, J. Chang, X. Li, and J. Wan, “Efficient oriented object detection via wavelet-based energy label reassignment and dual prediction strategy,”IEEE Transactions on Multimedia, 2025. 3

    2025

  41. [41]

    Category-aware dynamic label assignment with high-quality proposals for oriented object detection,

    M. Feng, H. Yu, X. Dang, and M. Zhou, “Category-aware dynamic label assignment with high-quality proposals for oriented object detection,” IEEE Transactions on Multimedia, 2025. 3

    2025

  42. [42]

    Ace: Anchor-free corner evolution for real-time arbitrarily-oriented object detection,

    P. Dai, S. Yao, Z. Li, S. Zhang, and X. Cao, “Ace: Anchor-free corner evolution for real-time arbitrarily-oriented object detection,”IEEE Transactions on Image Processing, vol. 31, pp. 4076–4089, 2022. 3

    2022

  43. [43]

    Instance-level orientation enhancement for horizontal box supervised oriented object detection in remote sensing images,

    Y . Xu, Z. Xu, H. Wang, Z. Wei, and Z. Wu, “Instance-level orientation enhancement for horizontal box supervised oriented object detection in remote sensing images,”IEEE Transactions on Image Processing, vol. 34, pp. 7613–7626, 2025. 3

    2025

  44. [44]

    Pra-det: Anchor-free oriented object detection with polar radius representation,

    M. Dang, G. Liu, H. Li, D. Wang, R. Pan, and Q. Wang, “Pra-det: Anchor-free oriented object detection with polar radius representation,” IEEE Transactions on Multimedia, vol. 27, pp. 145–157, 2024. 3

    2024

  45. [45]

    Bounding box vectorization for oriented object detection with tanimoto coefficient regression,

    L. Wang, Y . Zhan, W. Liu, B. Yu, and D. Tao, “Bounding box vectorization for oriented object detection with tanimoto coefficient regression,”IEEE transactions on multimedia, vol. 26, pp. 5181–5193,

  46. [46]

    Focal loss for dense object detection,

    T.-Y . Lin, P. Goyal, R. Girshick, K. He, and P. Doll ´ar, “Focal loss for dense object detection,” inProceedings of the IEEE international conference on computer vision, 2017, pp. 2980–2988. 6, 11

    2017

  47. [47]

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

    Z. Liu, L. Yuan, L. Weng, and Y . Yang, “A high resolution optical satellite image dataset for ship recognition and some new baselines,” inInternational conference on pattern recognition applications and methods, vol. 2. SciTePress, 2017, pp. 324–331. 6

    2017

  48. [48]

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

    G.-S. Xia, X. Bai, J. Ding, Z. Zhu, S. Belongie, J. Luo, M. Datcu, M. Pelillo, and L. Zhang, “Dota: A large-scale dataset for object detection in aerial images,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 3974–3983. 6

    2018

  49. [49]

    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. 6

    2022

  50. [50]

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

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

  51. [51]

    Redet: A rotation-equivariant detector for aerial object detection,

    J. Han, J. Ding, N. Xue, and G.-S. Xia, “Redet: A rotation-equivariant detector for aerial object detection,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 2786–

    2021

  52. [52]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778. 6

    2016

  53. [53]

    Mmrotate: A rotated object detection benchmark using pytorch,

    Y . Zhou, X. Yang, G. Zhang, J. Wang, Y . Liu, L. Hou, X. Jiang, X. Liu, J. Yan, C. Lyuet al., “Mmrotate: A rotated object detection benchmark using pytorch,” inProceedings of the 30th ACM International Confer- ence on Multimedia, 2022, pp. 7331–7334. 6

    2022

  54. [54]

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

    X. Yang, J. Yan, W. Liao, X. Yang, J. Tang, and T. He, “Scrdet++: Detecting small, cluttered and rotated objects via instance-level feature denoising and rotation loss smoothing,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 2, pp. 2384–2399, 2022. 10

    2022

  55. [55]

    Refined one-stage oriented object detection method for remote sensing images,

    L. Hou, K. Lu, and J. Xue, “Refined one-stage oriented object detection method for remote sensing images,”IEEE Transactions on Image Processing, vol. 31, pp. 1545–1558, 2022. 9

    2022

  56. [56]

    Fcos: Fully convolutional one- stage object detection,

    Z. Tian, C. Shen, H. Chen, and T. He, “Fcos: Fully convolutional one- stage object detection,” inProceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 9627–9636. 10, 11

    2019