Pith. sign in

REVIEW 4 major objections 5 minor 40 references

SARD: A YOLOv8-Based System for Solar Active Region Detection with SDO/HMI Magnetograms

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A YOLOv8-based detector finds solar active regions in full-disk magnetograms with 94% average precision and under one second per image.

desk verdict Useful, reproducible YOLOv8-based AR detector, but the paper's numbers don't add up and the statistical comparison is stacked; fixable with a careful revision. read the letter →

arxiv 2507.21866 v2 pith:Y466ABUP submitted 2025-07-29 astro-ph.SR

Jinhui Pan , Jiajia Liu , Shaofeng Fang , Rui Liu This is my paper · ORCID
classification astro-ph.SR
keywords solaractiveregionsobjectdetectionYOLOv8SDO/HMImagnetogramsdeeplearninglog-normaldistributioncycle24deformableconvolution
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

The paper sets out to show that a single-stage object detector built on YOLOv8, with three targeted modifications, can replace slow manual or image-processing identification of solar active regions. Trained on 26,531 labels from 6,975 SDO/HMI line-of-sight magnetograms spanning 2010 to 2019, and tested on temporally disjoint periods, the system reports an average precision of 94%, a precision of 0.89, and a recall of 0.88 while running in under one second per full-disk image. The same detector then feeds a statistical study in which the area and magnetic flux of active regions are both best described by log-normal distributions. If these results hold, researchers gain an automatic, fast way to build active-region catalogs for space-weather and solar-dynamo studies without waiting for manual annotation.

What carries the argument

The load-bearing object is the SARD network itself: YOLOv8s with deformable convolution (learnable offsets added to each sampling position, so the kernel can follow irregular active-region boundaries), an Efficient Channel Attention module (a fast 1D convolution that reweights feature channels based on global average pooling), and an extra detection head at 256x256 resolution for small regions. Supporting it is the dataset-generation pipeline: Otsu thresholding, morphological opening, region growing, morphological closing, then manual alignment of the resulting boxes to the NOAA catalog. The statistical claim is carried by fitting three candidate functions (log-normal, exponential, power law) to the area and flux histograms and testing the fits with a Kolmogorov-Smirnov test.

What would settle it

Build an independent test set of full-disk magnetograms in which active regions are manually outlined by several solar physicists according to a fixed magnetic-flux definition, deliberately excluding NOAA group boundaries and including weak and decayed regions; compute SARD's average precision at IoU>0.5 on that set. If it falls well below 0.94, the headline generalization fails; if the area and flux histograms from the independently segmented data are not log-normal, the statistical conclusion would have to be revised.

Watch

Extended reading notes

Core claim

On the paper's own terms, SARD is the discovery that a YOLOv8s base model extended with deformable convolution in its C2f blocks, an Efficient Channel Attention module, and an extra 256x256 small-object detection head outperforms every baseline tested: it reaches an average precision of 0.943, precision 0.89, recall 0.881, and F1 0.886 on the full test set assembled with a 14-day gap separating training and test images. Ablation experiments attribute AP gains of 0.2% to the extra head, 1.1% to ECA, and 0.9% to deformable convolution. When the detections are projected with a cylindrical equal-area projection and restricted to a central-meridian window, the area and magnetic flux distributions of 1,331 active regions are fit by log-normal curves, with Kolmogorov-Smirnov p-values above 0.05, while exponential and power-law fits are rejected. The paper reads this as support for the idea that magnetic flux elements fragment randomly in the convection zone before emerging as active regions.

Load-bearing premise

The load-bearing premise is that the NOAA active-region catalog, as reinterpreted through the paper's semi-automated labeling pipeline (thresholding, morphology, region growing, and manual adjustment), gives reliable and complete ground truth; if those labels are systematically biased, the reported 94% average precision and the log-normal statistics inherit that bias.

Editorial extensions

If this is right

  • SARD can process a full-disk magnetogram in under one second, making it practical to run continuously on SDO/HMI data streams and to reprocess the entire 2010-2019 archive cheaply.
  • Because the test sets share no time window with the training set, the 94% average precision indicates the detector generalizes across the solar cycle rather than memorizing daily patterns.
  • The log-normal area and flux distributions imply that a fragmentation process, rather than a single characteristic scale, governs active-region emergence.
  • The ablation ordering shows that the small-object detection head, channel attention, and deformable convolution each contribute additively, so further architecture work along these lines is a plausible route to higher recall on small, clustered regions.

Reading between the lines

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

  • One extension the paper leaves implicit: because low-activity periods trigger false positives on diffuse strong-field regions, a production version would likely need confidence calibration or a quiet-Sun suppression term before being trusted for flare-forecasting inputs.
  • A reader could test whether the log-normal result is an artifact of the NOAA-based bounding boxes, which draw human-defined boundaries around whole groups; an independent segmentation into individual magnetic concentrations might change the tail of the distribution.
  • The paper's comparison with YOLOv11 and YOLOv12 suggests newer object-detection backbones are not automatically better for this task; a natural follow-up is an architecture search over attention placement rather than scaling width and depth.
  • If SARD were retrained on vector magnetograms or on more recent data from solar cycle 25, the same pipeline would yield a near-real-time catalog that could be checked against flare occurrence to test whether detection confidence predicts eruption likelihood.
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 / 5 minor

Summary. This manuscript presents SARD, a YOLOv8-based object detection system for solar active regions (ARs) in SDO/HMI line-of-sight magnetograms. The authors construct a labeled dataset by applying thresholding, morphological operations, and region growing to 6,975 magnetograms from 2010 to 2019 and manually adjusting the automated labels to match NOAA/SolarMonitor AR annotations. They introduce three modifications to the YOLOv8 architecture (an additional small-object detection head, an ECA attention module, and deformable convolutions), train the model with a temporally split training/test protocol, and report precision around 0.89, recall around 0.88, and average precision around 0.94 on three test sets. They further fit log-normal, exponential, and power-law functions to the area and magnetic flux distributions of the detected ARs and conclude that log-normal distributions provide the best fit. The paper provides links to the dataset and code and includes an ablation study of the architectural changes.

Significance. If the reported performance is reproducible, SARD would be a practically useful tool for automated AR detection, and the public release of code and data is a positive contribution. The temporal split with a 14-day gap is a sensible attempt to prevent data leakage, and the ablation study gives partial credit for each architectural modification. However, the current text contains unresolved inconsistencies in dataset counts and metric values that prevent verification of the central 'no overlap' and 'AP as high as 94%' claims, and the statistical analysis of the log-normal result needs more careful fitting and model-comparison methodology. These issues are fixable but currently load-bearing.

major comments (4)
  1. [§2.2, Table 1, and §6] The dataset description is internally inconsistent. The abstract and §2.2 state 6,975 magnetograms and 26,531 labeled ARs, but the image counts in Table 1 sum to 5,063 + 311 + 790 + 699 = 6,863, and §6 states 19,098 labeled ARs. In addition, Test 2 is described as 'Entire 2010; 2011.1.1–2011.6.16' at a 12-hour cadence, which would contain roughly 1,000 images rather than the listed 790. Please provide a complete data manifest with per-set image identifiers and label counts, and reconcile all numbers. Without this, the no-overlap guarantee and the reported AP of 0.943 cannot be independently audited.
  2. [§4.2, Table 3, and Table 4] The performance numbers are inconsistent across the text and tables. The text reports overall precision, recall, F1, and AP of 0.895, 0.884, 0.889, and 0.946, while Table 3 and the 'Overall' row of Table 4 list 0.89, 0.881, 0.886, and 0.943; the T1 values also differ between the text (0.896/0.901/0.898/0.95) and Table 4 (0.893/0.898/0.896/0.947). Please use one consistent set of metrics, state the exact confidence threshold and IoU criterion used for each reported number, and clarify whether AP is computed at IoU = 0.5 only.
  3. [§5, Eq. (10)] The statistical analysis needs revision. The log-normal density in Eq. (10) is written with exp(−ln(x/μ)^2/(2σ^2)); with this parameterization μ is a scale parameter, not the expected value as the text states, so the fitted μ values are not directly comparable to the usual log-normal mean. The power-law fit fixes xmin to the median of the data, and the reported KS p-values are computed on the same data used to estimate the parameters; this is not a valid model comparison across distributions fitted over different ranges. Please provide the fitting algorithm, parameter uncertainties, and a fair model comparison (e.g., bootstrap or information criteria) before claiming that log-normal is the best fit. The unsupported assertion in §5 that excluding newly emerging ARs 'barely affected' the statistics should also be quantified.
  4. [§2.1 and §2.2] The ground-truth labels are produced by an automated pipeline and then manually adjusted to match NOAA/SolarMonitor AR annotations, but the paper does not quantify the frequency or extent of manual adjustment, nor does it compare the final labels with an independent AR catalog or estimate inter-annotator variability. Because both the AP values and the flux/area distributions are computed against these labels, the results are conditional on the NOAA AR definition. Please add a quantitative description of the labeling workflow and, ideally, a comparison of the final labels to the raw NOAA positions and sizes.
minor comments (5)
  1. [Figure 1 caption] The caption spells 'OSTU' where the text and standard usage write 'Otsu'; please correct the spelling.
  2. [§2.1 and §5] There are typographical errors: 'vales' should be 'values' in §2.1, and 'phtospheric' should be 'photospheric' in §5.
  3. [§4.2] The dates '12 October 2010' and '10 October 23' are garbled; AR 11856 belongs to 2013, and Figure 1 shows 12 October 2013, so the dates and figure references should be corrected.
  4. [Table 2 note] The note 'see Kingma (2014) for AdamW optimizer' is incorrect; AdamW is introduced by Loshchilov and Hutter (2019), while Kingma and Ba (2015) introduced Adam. Please update the citation.
  5. [§5] When reporting the KS-test results for the full dataset, the paper gives p-values for the flux distribution but does not give the corresponding p-values for the area distribution; please report them consistently for both distributions and all fitted models.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: SARD's AP is measured on held-out images against NOAA-based labels, and the log-normal fits are standard model comparisons; no claim reduces to its own input.

full rationale

The paper's central claims are a supervised object-detection evaluation and a distribution-fitting exercise, neither of which is circular. The detector is trained on labeled magnetograms and tested on disjoint time periods with a 14-day gap; the AP of 0.943 is computed against ground-truth labels that are independent of the model's own outputs, so the performance claim is not equivalent to a fitted input. The log-normal result is obtained by fitting three candidate functions (log-normal, exponential, power-law) to the same data and comparing KS-test p-values; this is a standard model-comparison procedure, not a prediction forced by construction. The only self-citations (Liu et al. 2012, 2019) appear in the introduction as background context and are not load-bearing. The paper does contain internal numeric inconsistencies (e.g., 6,975 vs. 6,863 magnetograms; 26,531 vs. 19,098 labeled ARs; AP values 0.943 vs. 0.946), but these are data-accounting and auditability concerns, not evidence of circular reasoning. No step in the derivation chain defines a target quantity in terms of itself, disguises a fit as a prediction, or imports a uniqueness claim from the authors' prior work.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The central detection claim rests on the labeling pipeline parameters (thresholds, structuring element, saturation limit) and on the reliability of the NOAA-based ground truth. The statistical claim rests further on the fitted distribution parameters and on the domain approximation that LoS field approximates the vertical component.

free parameters (7)
  • Log-normal mu and sigma for flux distribution = Not reported in text; shown in Figure 6
    Maximum-likelihood parameters for the flux histogram; the log-normal conclusion depends on these values.
  • Log-normal mu and sigma for area distribution = Not reported in text; shown in Figure 6
    Maximum-likelihood parameters for the area histogram; the log-normal conclusion depends on these values.
  • Exponential mu for flux and area fits = Not reported in text; shown in Figure 6
    Used as a comparison model; deliberately not the chosen model.
  • Power-law alpha and xmin = xmin fixed to median; alpha fitted
    xmin fixed to the median value of the data, a choice that severely constrains the power-law fit and effectively predisposes the test toward rejecting it.
  • Region-growing intensity threshold = 0.05
    Hand-chosen threshold in the labeling pipeline; affects the boundaries of the ground-truth ARs.
  • Morphological structuring element size = 11x11
    Hand-chosen square element for opening and closing operations during label preparation.
  • Magnetogram saturation limit = 650 G
    Hand-chosen range for image saturation in pre-processing, affecting the contrast and the labeling result.
assumptions (4)
  • domain assumption Line-of-sight magnetic field approximates the vertical component for AR flux estimation
    Stated in Section 5: 'we took the LoS component of the photospheric magnetic field as a good approximation of the vertical component'. This affects the flux distribution result.
  • domain assumption NOAA AR catalog provides complete and accurate AR boundaries for ground truth
    The authors manually adjusted automated labels to match NOAA ARs from SolarMonitor (Section 2.1), making NOAA the de facto ground truth for all performance metrics and for the statistical analysis.
  • domain assumption Bounding boxes can adequately represent solar ARs for detection and statistics
    The YOLOv8 framework outputs bounding boxes; ARs with irregular shapes are approximated by rectangles (Figure 2 and Section 3). This is inherent to the chosen method.
  • standard math KS test with p > 0.05 indicates a good fit
    Used in Section 5 to select log-normal over exponential and power-law. The KS test's validity depends on the fits being computed from the same data without prior selection; this is an unstated statistical assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SARD: A YOLOv8-Based System for Solar Active Region Detection with SDO/HMI Magnetograms." pith.science (2026). https://pith.science/paper/Y466ABUP

@misc{pith2026250721866,
  author       = {Pith},
  title        = {Pith review of: SARD: A YOLOv8-Based System for Solar Active Region Detection with SDO/HMI Magnetograms},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y466ABUP}},
  note         = {Machine review of arXiv:2507.21866}
}
read the original abstract

Solar active regions are where sunspots are located and photospheric magnetic fluxes are concentrated, therefore being the sources of energetic eruptions in the solar atmosphere. The detection and statistics of solar active regions have been forefront topics in solar physics. In this study, we developed a solar active region detector (SARD) based on the advanced object detection model YOLOv8. First, we applied image processing techniques including thresholding and morphological operations to 6975 line-of-sight magnetograms from 2010 to 2019 at a cadence of 12~h, obtained by the Helioseismic and Magnetic Imager onboard the Solar Dynamic Observatory. With manual refinement, we labeled 26531 active regions in the dataset for further training and test with the detection model. Without any overlap between the training and test sets, the superior performance of SARD is demonstrated by an average precision rate as high as 94\%. We then performed a statistical analysis on the area and magnetic flux of the detected active regions, both of which yield log-normal distributions. This result sheds light on the underlying complexity and multi-scale nature of solar active regions.

Figures

Figures reproduced from arXiv: 2507.21866 by the authors.

Figure 1
Figure 1. Flow chart of the image pre-processing during dataset preparation. Panel (a) shows the original LoS magnetogram of at 00:00:00 TAI on 12 October 2013. Panel (b) displays the result after OSTU thresholding. Panel (c) illustrates the outcome of the morphological opening operation, and Panel (d) shows the result of the region-growing method. Panel (e) presents the result of the morphological closing operation, while Pa… view at source ↗
Figure 2
Figure 2. Configuration of SARD. The details of substructures are also illustrated. volutional modules and multiple bottlenecks. However, the common convolution kernels in this module restrict the receptive field of the network, and only capture local feature information (Wei et al. 2022). Deformable convolutional networks (DCN) introduce learnable offsets, enabling convolutional neural networks to adaptively adjust their rec… view at source ↗
Figure 3
Figure 3. Configuration of Efficient Channel Attention (ECA) mechanism. Unlike traditional channel attention mechanisms, ECA avoids the complexity of dimensionality reduction and captures local cross-channel interactions using a fast 1D convolution ( [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The P-R curves of YOLOv8n, YOLOv8s, YOLOv8m, YOLOv8l, and SARD on the entire test set. P and R are calculated at different confidence thresholds, and the area under the curve is the AP. The dots indicate the final reported values of precision and recall. With precision…
Figure 5
Figure 5. Figure 5: Examples of ARs detected by the SARD using LoS magnetograms. NOAA ARs are artificially marked by blue boxes with the AR numbers annotated. The red bounding boxes are given by the SARD. of this experiment are presented in [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: The flux and area distributions in log-log scales. The histograms represent the distributions of the entire dataset, while the scatter points represent the test set, with black and yellow colors indicating the ground truth (GT) and the SARD-detected results, respective…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 36 canonical work pages

  1. [1]

    Solar Physics 262,

    Aschwanden, M.J.: 2010, Image processing techniques and feature recognition in solar physics. Solar Physics 262,

    work page 2010

  2. [3]

    DOI. ADS. SOLA: sola_example_6.tex; 12 September 2025; 4:49; p. 18 Solar Active Region Detection Portier-Fozzani, F., Vandame, B., Bijaoui, A., Maucherat, A., Team, E.: 2001, A Multiscale Vision Model applied to analyze EIT images of the solar corona. Solar Physics 201,

    work page 2025

  3. [7]

    Solar Physics 250,

    Curto, J., Blanca, M., Mart ´ ınez, E.: 2008, Automatic sunspots detection on full-disk solar images using mathematical morphology. Solar Physics 250,

    work page 2008

  4. [15]

    arXiv preprint arXiv:1412.6980

    Kingma, D.P.: 2014, Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980. Li, C., Li, L., Jiang, H., Weng, K., Geng, Y., Li, L., Ke, Z., Li, Q., Cheng, M., Nie, W., et al.: 2022, YOLOv6: A single-stage object detection framework for industrial applications. arXiv preprint arXiv:2209.02976. Lin, T.-Y., Doll´ ar, P., Girshick, R., He, ...

    arXiv 2014

  5. [23]

    Pesnell, W.D., Thompson, B.J., Chamberlin, P.C.: 2012, The Solar Dynamics Observatory (SDO). Sol. Phys. 275,

    work page 2012

  6. [33]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition,

    Wang, C.-Y., Bochkovskiy, A., Liao, H.-Y.M.: 2023, 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,

    work page 2023

  7. [36]

    Computers and Electronics in Agriculture 193, 106653

    Wei, H., Xu, E., Zhang, J., Meng, Y., Wei, J., Dong, Z., Li, Z.: 2022, BushNet: Effective semantic segmentation of bush in large-scale point clouds. Computers and Electronics in Agriculture 193, 106653. DOI. https://www.sciencedirect.com/science/article/pii/S0168169921006700. Wolf: 1861, Abstract of his latest Results. Monthly Notices of the Royal Astrono...

    work page 2022

  8. [55]

    Solar Physics 125,

    McIntosh, P.S.: 1990, The classification of sunspot groups. Solar Physics 125,

    work page 1990

Show all 40 references
  1. [59]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition,

    Liu, S., Qi, L., Qin, H., Shi, J., Jia, J.: 2018, Path aggregation network for instance segmen- tation. In: Proceedings of the IEEE conference on computer vision and pattern recognition,

    2018

  2. [67]

    In: Proceedings of the IEEE/CVF international conference on computer vision,

    Duan, K., Bai, S., Xie, L., Qi, H., Huang, Q., Tian, Q.: 2019, Centernet: Keypoint triplets for object detection. In: Proceedings of the IEEE/CVF international conference on computer vision,

    2019

  3. [75]

    Astronomy & Astrophysics 449,

    Thompson, W.: 2006, Coordinate systems for solar image data. Astronomy & Astrophysics 449,

    2006

  4. [77]

    https://doi.org/10.1093/mnras/21.3.77

    DOI. https://doi.org/10.1093/mnras/21.3.77. Xie, S., Girshick, R., Doll´ ar, P., Tu, Z., He, K.: 2017, Aggregated residual transformations for deep neural networks. In: Proceedings of the IEEE conference on computer vision and pattern recognition,

    2017 doi

  5. [93]

    Automatica 11,

    Otsu, N., et al.: 1975, A threshold selection method from gray-level histograms. Automatica 11,

    1975

  6. [121]

    Solar Physics 252,

    Ireland, J., Young, C., McAteer, R., Whelan, C., Hewett, R., Gallagher, P.: 2008, Multiresolu- tion analysis of active region magnetic structure and its correlation with the Mount Wilson classification and flaring activity. Solar Physics 252,

    2008

  7. [160]

    Solar Physics 262,

    Biskri, S., Antoine, J.-P., Inhester, B., Mekideche, F.: 2010, Extraction of solar coronal magnetic loops with the directional 2D morlet wavelet transform. Solar Physics 262,

    2010

  8. [207]

    DOI. ADS. Schrijver, C.J., Title, A.M., Hagenaar, H.J., Shine, R.A.: 1997, Modeling the distribution of magnetic fluxes in field concentrations in a solar active region. Solar Physics 175,

    1997

  9. [235]

    SOLA: sola_example_6.tex; 12 September 2025; 4:49; p. 17 J. Pan et al. Baek, J.-H., Kim, S., Choi, S., Park, J., Kim, J., Jo, W., Kim, D.: 2021, Solar event detection using deep-learning-based object detection methods. Solar Physics 296,

    2025

  10. [251]

    Solar physics 149,

    Ograpishvili, N.: 1994, Distribution of active regions on the solar surface estimated by statistical methods. Solar physics 149,

    1994

  11. [271]

    Electronics 10,

    Quan, L., Xu, L., Li, L., Wang, H., Huang, X.: 2021, Solar active region detection using deep learning. Electronics 10,

    2021

  12. [329]

    Solar physics 91,

    Tang, F., Howard, R., Adkins, J.M.: 1984, A statistical study of active regions 1967–1981. Solar physics 91,

    1984

  13. [373]

    arXiv preprint arXiv:2004.10934

    Bochkovskiy, A., Wang, C.-Y., Liao, H.-Y.M.: 2020, Yolov4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934. Bogdan, T., Gilman, P.A., Lerche, I., Howard, R.: 1988, Distribution of sunspot umbral areas- 1917-1982. Astrophysical Journal, Part 1 (I...

    2020 arXiv

  14. [377]

    SOLA: sola_example_6.tex; 12 September 2025; 4:49; p. 19

    2025

  15. [390]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 11534

    Wang, Q., Wu, B., Zhu, P., Li, P., Zuo, W., Hu, Q.: 2020b, ECA-Net: Efficient channel attention for deep convolutional neural networks. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 11534. Wang, R., Jiang, J., Luo, Y.: 2023, Toward a Li...

    2023

  16. [411]

    In: Proceedings of the IEEE international conference on computer vision,

    Dai, J., Qi, H., Xiong, Y., Li, Y., Zhang, G., Hu, H., Wei, Y.: 2017, Deformable convolutional networks. In: Proceedings of the IEEE international conference on computer vision,

    2017

  17. [451]

    Solar Physics 289,

    Caballero, C., Aranda, M.: 2014, Automatic tracking of active regions and detection of solar flares in solar EUV images. Solar Physics 289,

    2014

  18. [580]

    Solar Physics 299,

    Gong, L., Yang, Y., Feng, S., Dai, W., Liang, B., Xiong, J.: 2024, Solar active regions detection and tracking based on deep learning. Solar Physics 299,

    2024

  19. [764]

    Dang, T., Li, X., Luo, B., Li, R., Zhang, B., Pham, K., Ren, D., Chen, X., Lei, J., Wang, Y.: 2022, Unveiling the space weather during the Starlink satellites destruction event on 4 February

    2022

  20. [779]

    Advances in neural information processing systems28

    Ren, S., He, K., Girshick, R., Sun, J.: 2015, Faster r-cnn: Towards real-time object detection with region proposal networks. Advances in neural information processing systems28. Scherrer, P.H., Schou, J., Bush, R.I., Kosovichev, A.G., Bogart, R.S., Hoeksema, J.T., Liu, Y., Du...

    2015

  21. [791]

    van Driel-Gesztelyi, L., Green, L.M.: 2015, Evolution of active regions.Living Reviews in Solar Physics 12,

    2015

  22. [1006]

    Detection, North–South asymmetry and area distribution

    Zharkov, S., Zharkova, V., Ipson, S.: 2005, Statistical properties of sunspots in 1996–2004: I. Detection, North–South asymmetry and area distribution. Solar Physics 228,

    2005

  23. [1492]

    The Astrophysical Journal 723,

    Zhang, J., Wang, Y., Liu, Y.: 2010, Statistical properties of solar active regions obtained from an automatic detection system and the computational biases. The Astrophysical Journal 723,

    2010

  24. [1643]

    Space weather

    Colak, T., Qahwaji, R.: 2009, Automated Solar Activity Prediction: A hybrid computer plat- form using machine learning and solar imaging for automated prediction of solar flares. Space weather

    2009

  25. [2022]

    Das, T., Das Gupta, M.: 1982, Distribution of sunspots according to their magnetic fluxes/Research note

    Space weather 20, e2022SW003152. Das, T., Das Gupta, M.: 1982, Distribution of sunspots according to their magnetic fluxes/Research note. Solar Physics, vol. 78, May 1982, p. 67-70.78,

    1982

  26. [2117]

    Nature Communications 10,

    Liu, J., Nelson, C.J., Snow, B., Wang, Y., Erd´ elyi, R.: 2019, Evidence of ubiquitous Alfv´ en pulses transporting energy from the photosphere to the upper chromosphere. Nature Communications 10,

    2019

  27. [2284]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition,

    Redmon, J., Farhadi, A.: 2017, YOLO9000: better, faster, stronger. In: Proceedings of the IEEE conference on computer vision and pattern recognition,

    2017

  28. [3504]

    Liu, R., Kliem, B., T¨ or¨ ok, T., Liu, C., Titov, V.S., Lionello, R., Linker, J.A., Wang, H.: 2012, Slow rise and partial eruption of a double-decker filament. I. Observations and interpretation. The Astrophysical Journal756,

    2012

  29. [6569]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition,

    Girshick, R., Donahue, J., Darrell, T., Malik, J.: 2014, Rich feature hierarchies for accurate object detection and semantic segmentation. In: Proceedings of the IEEE conference on computer vision and pattern recognition,

    2014

  30. [7263]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition,

    Redmon, J., Divvala, S., Girshick, R., Farhadi, A.: 2016, You only look once: Unified, real-time object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition,

    2016

  31. [7464]

    arXiv preprint arXiv:2211.04800

    Wang, C.-Y., Liao, H.-Y.M., Yeh, I.-H.: 2022, Designing network design strategies through gradient path analysis. arXiv preprint arXiv:2211.04800. Wang, C.-Y., Liao, H.-Y.M., Wu, Y.-H., Chen, P.-Y., Hsieh, J.-W., Yeh, I.-H.: 2020a, CSP- Net: A new backbone that can enhance lea...

    2022 arXiv

  32. [8759]

    Solar Physics 228,

    McAteer, R.J., Gallagher, P.T., Ireland, J., Young, C.A.: 2005, Automated boundary- extraction and region-growing techniques applied to solar magnetograms. Solar Physics 228,

    2005

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.