Pith. sign in

REVIEW 4 major objections 5 minor 76 references

Towards Large Scale Geostatistical Methane Monitoring with Part-based Object Detection

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 ↗ Review JSON

Pith's one-line read A part-based probabilistic post-processing step, combined with iterative hard negative mining, lets a detector trained on only 163 annotated sites find bio-digesters across whole regions of France and support bottom-up estimates of…

desk verdict A genuinely new bio-digester dataset and a sensible detection pipeline, but the 'nationwide' and methane-estimate claims outrun the evidence; worth refereeing with a required revision of Section 5.3 and the evaluation protocol. read the letter →

arxiv 2507.18513 v1 pith:4ZBCQOIR submitted 2025-07-24 cs.CV

Adhemar de Senneville , Xavier Bou , Thibaud Ehret , Rafael Grompone , Jean Louis Bonne , Nicolas Dumelie , Thomas Lauvaux , Gabriele Facciolo This is my paper · ORCID
classification cs.CV
keywords methanemonitoringbio-digesterdetectionpart-basedobjectremotesensinghardnegativemininginventoryingsatelliteimagerygeostatisticalestimation
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 is trying to establish that a detector trained on only 163 annotated satellite images can find rare methane-producing bio-digesters across whole regions of France, provided the raw detections are filtered by a statistical check for the site's characteristic parts and the worst false positives are mined and fed back into training. If this works, it offers an imagery-based way to keep methane inventories of small, diffuse emitters current, since these facilities are too small for current satellite methane sensors and are otherwise tracked through self-reporting. The paper shows the method finds sites missing from existing databases (about 25% of its Grand Est detections were previously unknown), that it generalizes to other regions with precision near 80% in Marne and 71% in Bretagne, and that tank area measured from the inventory predicts reported power production well enough for aggregated estimates. The central claim is therefore that small-source methane inventories can be built bottom-up from satellite imagery at national scale from a minimal labeling effort.

What carries the argument

The load-bearing object is the part-based probabilistic score of Equation (2): a bio-digester detection is confirmed by counting digestion tanks and biomass piles inside its bounding box, where each sub-detection is an independent Bernoulli trial and the probability of a count is a Poisson-binomial sum. The final score $p(D | p_t, p_p, p_b) = p_b \sum_{N_t,N_p} p(N_t | p_t) p(N_p | p_p) p(D | N_t, N_p)$ combines the raw site confidence with the probability that a real bio-digester would have that many parts, using a prior histogram estimated from the training set. This turns the raw detector score into a posterior-like measure that suppresses look-alike industrial structures. The second mechanism is the iterative hard-negative mining loop: after each deployment, the top-K most confident false detections are human-verified and added to the training set as background tiles, which is what raises tank AP50 from 0.16 to 0.86.

What would settle it

Take a region with an independently complete register of bio-digesters and their per-site tank and pile counts, run the full pipeline without retraining, and compare the detected counts of tanks and piles at each matched site to the registered ones; if real sites are rejected or precision collapses because the per-site count distribution differs from the Grand Est histogram, the central transferability claim fails.

Watch

Extended reading notes

Core claim

The central claim is that a part-based probabilistic post-processing step, defined by Equation (2), removes most of the false alarms a conventional detector produces at the low confidence thresholds needed for high recall at scale. For each candidate bio-digester bounding box, the tank and pile detections inside it are treated as independent Bernoulli trials, giving a probability for the observed counts; this is multiplied by the raw site confidence and by a histogram-based prior $p(D | N_t, N_p)$ learned from the 163 training sites. Iterating this pipeline and injecting the most confident false detections as hard negatives raises mAP50 from 0.26 to 0.59 across three iterations. The resulting model, applied to the Grand Est region, Marne, and Bretagne, identifies new sites absent from official databases, and the inventory's summed tank areas in Bretagne yield a linear production estimate with $r^2 = 0.332$.

Load-bearing premise

The load-bearing premise is that the probability of a site being a bio-digester given its number of tanks and piles, estimated from the 163 Grand Est training sites, transfers to other regions; if that count prior is wrong elsewhere, the part-based filter either keeps false alarms or throws away real sites.

Editorial extensions

If this is right

  • A model trained on one region can be deployed over unannotated regions to produce or update bio-digester inventories, finding sites that official databases omit.
  • Aggregated methane production of an area can be estimated bottom-up from the inventory by regressing power on total tank area, giving a check on self-reported inventories.
  • SPOT 1.5 m satellite data, which underperforms aerial imagery only slightly, makes repeated large-scale scans feasible and opens the door to temporal tracking of facility construction and production growth.
  • The large-scale dataset released with the paper, including hard negatives and new detections, provides a starting resource for further methane-inventorying work.

Reading between the lines

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

  • The same part-based recipe could be applied to other composite small emitters with characteristic sub-structures, such as wastewater treatment plants or landfill gas facilities, provided the per-site part-count prior is re-estimated for each object class and region.
  • The Bretagne recall drop (53.6% versus 83.9% in Grand Est) suggests the tank/pile count prior learned in Grand Est does not fully transfer; estimating the prior per region from a handful of known sites, or adapting it online, would be a direct improvement to test.
  • Because the power regression explains only 33% of the variance, aggregated estimates should be validated against independent regional production statistics (e.g., grid injection data) across several departments before relying on them for emission factors.
  • A stronger temporal use would pair the inventory with hyperspectral plume detections, which are limited to large emitters; the inventory's per-site tank area could serve as a prior for allocating small-source emissions at sub-regional scale.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The manuscript presents a pipeline for detecting agricultural bio-digester sites in aerial and satellite imagery from a small annotated seed set. It trains a conventional object detector on 163 Grand Est sites, then applies a part-based probabilistic post-processing step (Eq. 2) that combines detector scores for tanks and piles with a count prior, and iteratively mines hard negatives. The authors report precision/recall on Grand Est, Marne, and Bretagne, study the impact of image resolution and source (BD ORTHO, SPOT, Sentinel-2), and finally fit a linear regression from detected tank area to reported power production in Bretagne to support aggregated methane estimates. A dataset, code, and detected locations are promised for release.

Significance. The detection pipeline addresses a real and timely problem, and the paper is transparent about moderate out-of-region recall (53.6% in Bretagne, Table 2). The part-based score is an interesting, low-cost way to suppress false alarms when only a handful of annotated sites are available, and iterative hard-negative mining shows clear gains (mAP50 rising from 0.26 to 0.59 in Table 6). If the regression step were validated, the work would provide a useful bridge from imagery inventories to environmental accounting. However, the paper's central 'geostatistical methane monitoring' claim currently rests on an unvalidated in-sample fit, so the significance of the methane-estimation part is not yet established.

major comments (4)
  1. [§5.3 (Fig. 8)] The power-production model is fit to the Bretagne data and then used to produce an aggregated estimate for the same data; no held-out sites, cross-validation, prediction intervals, or aggregation-error analysis is provided. Because the abstract and title promise 'geostatistical estimates of the quantity of methane produced,' this in-sample r²=0.332 is load-bearing and cannot support the claim as written. I ask for a validation protocol (e.g., leave-one-site-out, a held-out region, or a bound on aggregate estimation error) or a substantial reframing of the methane-estimation contribution.
  2. [§5, Metrics] The true-positive distance threshold of 200 m is justified as 'the maximum offset observed among true positives in our data,' i.e., it is selected on the same evaluation data used to report Table 2. This can inflate reported precision and recall; a threshold chosen from the test data should be accompanied by a sensitivity analysis across distances (e.g., 50–300 m) or by a pre-registered physical rationale for site-matching distance.
  3. [§5, Generalization to other locations (Table 2, Fig. 3)] The Marne test set lies inside the Grand Est region used to collect training and validation images (Figure 3), so it is not an independent out-of-region test; the only genuinely out-of-region evaluation is Bretagne, where recall drops to 53.6%. The claim of 'robustly detects bio-digesters nationwide' is therefore stronger than the evidence. Please either restrict the generalization claim to the supported regions or add a true out-of-region test set.
  4. [§4, Eq. (2) and Fig. 5] The prior p(D|Nt,Np) is estimated from the 163-site Grand Est training histogram (Figure 5) and applied globally in Eq. (2). If the tank/pile count distribution differs across regions, the post-processing could suppress true detections or admit false alarms. The Bretagne recall drop is consistent with this risk. A sensitivity analysis of Eq. (2) to the prior, or a re-estimation of the prior on Bretagne counts, would make the nationwide claim defensible.
minor comments (5)
  1. [References] References [51] and [54] appear to describe the same dataset, but [54] is attributed to 'Scientific Data Curation Team' rather than the original authors and has a title mismatch. Please correct the citation.
  2. [§5, Metrics] The sentence 'This threshold corresponds to the maximum offset observed among true positives in our data' uses 'our data' ambiguously; specify whether this is training, validation, or test data.
  3. [Table 6] The Iteration column starts at 0; the text says three iterations including the first initial training, which is consistent but could be clarified for readers.
  4. [Figure 1] The caption says the part-based detector 'reliably identifies bio-digester sites at scale' while Table 2 reports 53.6% recall in Bretagne; consider softening the caption or qualifying the claim to the training region.
  5. [Throughout] Minor language issues: 'in the order of tons per hour' should be 'on the order of tons per hour', and the article sometimes uses 'a bio-digester' as an adjective ('a bio-digester site'), which is acceptable but slightly awkward.

Circularity Check

1 steps flagged · score 5.0 of 10

The methane-estimation bridge is an in-sample regression presented as a prediction; the detector itself is independently evaluated.

  1. fitted input called prediction [Section 5.3, Figure 8]
    "We build a linear regression model, illustrated in Figure 8, which predicts power production based on the overall detected tank area in a site (a feature linked to the scale of the facility). While unsurprisingly the model has a large predictive error, it is able to provide reasonable aggregated estimates over a large number of observations. Overall, the power production estimation model predicts 33% of the power variability."

    The regression is fit on the Bretagne sites and the reported r²=0.332 is the in-sample coefficient of determination of that same fit; it is not a held-out prediction score. Yet Section 5.3 and the abstract present this as 'predicting' power production and as supporting aggregated methane estimates. No cross-validation, no train/test split, no prediction interval, and no aggregation-error analysis is provided, so the 'prediction' statistic reduces by construction to the fit's R² and the aggregate estimate is an in-sample extrapolation of the fitted line.

full rationale

The detection pipeline is largely self-contained: the LSKNet detector is fine-tuned on independently annotated Grand Est sites, tested on Marne and Bretagne, and compared against pre-trained baselines, so the detection results do not reduce to their training inputs. The part-based probability Eq. (2) uses a prior histogram estimated from the training data, but this is a model component evaluated on separate validation/test images, not a prediction of the same data. No load-bearing uniqueness theorem or self-citation chain is present. The one substantial circularity is in the methane-estimation claim: Section 5.3 fits a linear regression of power on tank area on the Bretagne sites and then describes the in-sample R²=0.332 as the model 'predicts 33% of the power variability', and uses the same fit to produce aggregated kW estimates. Since no held-out validation, prediction intervals, or aggregation-error analysis are provided, this step is a fitted-value statistic renamed as a prediction. This is load-bearing because the title and abstract promise geostatistical methane estimates, not just detections.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The method introduces no new physical entities. The free parameters are the empirical prior on tank/pile counts, the 200 m evaluation threshold, the power regression coefficients, and the iterative-training hyperparameters. The main axioms are modeling assumptions about independence, prior transfer, calibration, and the tank-area to power proxy.

free parameters (5)
  • Empirical prior p(D|Nt,Np) over tank and pile counts per site = Histogram from 163 training sites (Figure 5)
    Used in Eq. (2) to compute the part-based detection score; estimated from the small training set and assumed to generalize to unseen regions.
  • True-positive distance threshold = 200 m
    Chosen as the maximum offset among true positives in the data; directly defines what counts as a detection in APdist and all reported precision/recall.
  • Power production linear regression coefficients = y = 0.19x (proportional), y = 0.20x + 11.5 kW (affine)
    Fit to tank area vs power in Bretagne data; used to produce aggregated production estimates, but evaluated on the same fit data (r^2 = 0.332).
  • Number of hard negatives per iteration and iteration count = K = 100, 3 iterations
    Hyperparameters of the iterative refinement; chosen by the authors without ablation.
  • Detection confidence threshold for large-scale inference = Not specified in the text
    The paper refers to a 'predefined confidence threshold' for extracting detections but does not report its value, making large-scale counts hard to reproduce exactly.
assumptions (5)
  • ad hoc to paper Part detections are independent Bernoulli trials with probabilities equal to detector scores
    Stated in Eq. (1) and its surrounding text; no justification for independence between tank and pile detections, which are physically co-located and often appear together.
  • domain assumption Prior distribution of tank and pile counts in unseen regions matches the training-set histogram
    The part-based score uses p(D|Nt,Np) estimated from 163 Grand Est sites; the Bretagne recall drop (53.6%) suggests this prior does not transfer perfectly.
  • domain assumption Detector class probabilities are calibrated, usable as exact probabilities
    Eq. (2) treats pb, pt, pp as probabilities and combines them via the law of total probability; object detector confidence scores are typically not well calibrated.
  • domain assumption Tank area detected in imagery is a valid linear proxy for installed power
    Section 5.3 regresses production power on tank area; this assumes consistent relationship between visible infrastructure and capacity across regions and vintages.
  • domain assumption BD ORTHO/SPOT annotations transferred across image sources remain temporally consistent
    The paper itself notes annotations transferred from IGN data introduce temporal inconsistencies (Section 5.1), so the SPOT evaluation relies on this assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Large Scale Geostatistical Methane Monitoring with Part-based Object Detection." pith.science (2026). https://pith.science/paper/4ZBCQOIR

@misc{pith2026250718513,
  author       = {Pith},
  title        = {Pith review of: Towards Large Scale Geostatistical Methane Monitoring with Part-based Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4ZBCQOIR}},
  note         = {Machine review of arXiv:2507.18513}
}
read the original abstract

Object detection is one of the main applications of computer vision in remote sensing imagery. Despite its increasing availability, the sheer volume of remote sensing data poses a challenge when detecting rare objects across large geographic areas. Paradoxically, this common challenge is crucial to many applications, such as estimating environmental impact of certain human activities at scale. In this paper, we propose to address the problem by investigating the methane production and emissions of bio-digesters in France. We first introduce a novel dataset containing bio-digesters, with small training and validation sets, and a large test set with a high imbalance towards observations without objects since such sites are rare. We develop a part-based method that considers essential bio-digester sub-elements to boost initial detections. To this end, we apply our method to new, unseen regions to build an inventory of bio-digesters. We then compute geostatistical estimates of the quantity of methane produced that can be attributed to these infrastructures in a given area at a given time.

Figures

Figures reproduced from arXiv: 2507.18513 by the authors.

Figure 1
Figure 1. Overview of our results in the French Grand Est region with the number of detected bio-digester sites in each department in 2023. We use our model to detect unknown bio-digester sites in large areas. On the right, we show (a) some examples of annotated bio-digester sites (from the validation set) with their sub-elements. (b) Shows predictions from our model, even with detection errors and a small training set, the p… view at source ↗
Figure 3
Figure 3. Map of the Grand Est region of France. It includes the test set locations in red only present in the department of Marne. The train-val set locations in green are located in the rest of the Grand Est region. Dataset Number of Annotated vs α Images Background Tiles Training 326 163 vs 163 50% Validation 440 40 vs 400 9.1% Test - Marne 5096 27 vs 5069 0.53% [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 2
Figure 2. Illustration of aerial and satellite datasets. with their annotations. (a) and (b) correspond to BD ORTHO images resam￾pled at 0.5 and 1.5m per pixel, respectively. Sentinel-2 (c) and SPOT (d) images are shown on the bottom. cated segmentation annotations. For each bio-digester, we annotated three different classes: the whole bio-digester in￾stallation, the (anaerobic) digestion tanks, and biomass piles (feedstock s… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: General diagram of the proposed approach. The initially annotated dataset is used to train a conventional object detector network. Then, we apply a part-based statistical method to boost detection performance, making large-scale detection bearable. Subse￾quently, an it…
Figure 5
Figure 5. Figure 5: Histogram of per-site number of digestion tanks and biomass piles in the train set. do so, we consider two approaches: (a) modeling with a bi￾variate Poisson distribution, and (b) using the empirical his￾togram under the assumption of independence (as the two variables…
Figure 6
Figure 6. Figure 6: Precision-Recall curves for different detection meth￾ods over all the Grand Est region. The Baseline method (blue) performs the worst, while the probabilistic part-based method (or￾ange) significantly boosts the performance. Region In-Database Overall TP GT Recall Corr…
Figure 7
Figure 7. Figure 7: Examples of hard negatives found during the iterative process. They were initially detected as false positives, as they do not contain bio-digester sites. Iteration Known Database Size New Detection Hard Negatives 0 203 - 0 1 203 149 100 2 352 205 200 [PITH_FULL_IMAGE…
Figure 9
Figure 9. Figure 9: Visualization of annotation methods on piles [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 10
Figure 10. Figure 10: Hard negative examples across three training iterations. At iteration 0, samples are randomly drawn from the Grand-Est region. 2 [PITH_FULL_IMAGE:figures/full_fig_p013_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

76 extracted references · 67 canonical work pages

  1. [1]

    Carte & chiffres cl ´es - la fili `ere m´ethanisation en bretagne et pays de la loire

    Association AILE. Carte & chiffres cl ´es - la fili `ere m´ethanisation en bretagne et pays de la loire. https: //aile.asso.fr/biogaz/la- filiere/carte- chiffres-cles-methanisation/, 2025. Accessed: June 12, 2025

    work page 2025

  2. [2]

    Reduc- ing false alarms in video surveillance by deep feature statis- tical modeling, 2023

    Xavier Bou, Aitor Artola, Thibaud Ehret, Gabriele Facciolo, Jean-Michel Morel, and Rafael Grompone von Gioi. Reduc- ing false alarms in video surveillance by deep feature statis- tical modeling, 2023

    work page 2023

  3. [3]

    Exploring robust features for few-shot object detection in satellite imagery, 2024

    Xavier Bou, Gabriele Facciolo, Rafael Grompone von Gioi, Jean-Michel Morel, and Thibaud Ehret. Exploring robust features for few-shot object detection in satellite imagery, 2024

    work page 2024

  4. [4]

    Structure tensor rep- resentation for robust oriented object detection, 2024

    Xavier Bou, Gabriele Facciolo, Rafael Grompone von Gioi, Jean-Michel Morel, and Thibaud Ehret. Structure tensor rep- resentation for robust oriented object detection, 2024

    work page 2024

  5. [5]

    A survey on object detec- tion in optical remote sensing images

    Gong Cheng and Junwei Han. A survey on object detec- tion in optical remote sensing images. ISPRS journal of pho- togrammetry and remote sensing, 117:11–28, 2016

    work page 2016

  6. [6]

    Learning rotation-invariant convolutional neural networks for object detection in vhr optical remote sensing images

    Gong Cheng, Peicheng Zhou, and Junwei Han. Learning rotation-invariant convolutional neural networks for object detection in vhr optical remote sensing images. IEEE Trans- actions on Geoscience and Remote Sensing , 54(12):7405– 7415, 2016

    work page 2016

  7. [7]

    MMYOLO: OpenMMLab YOLO series toolbox and benchmark

    MMYOLO Contributors. MMYOLO: OpenMMLab YOLO series toolbox and benchmark. https://github.com/ open-mmlab/mmyolo, 2022

    work page 2022

  8. [8]

    Cusworth, Andrew K

    Daniel H. Cusworth, Andrew K. Thorpe, Alana K. Ayasse, David Stepp, Joseph Heckler, Gregory P. Asner, Charles E. Miller, Vineet Yadav, John W. Chapman, Michael L. East- wood, et al. Strong methane point sources contribute a disproportionate fraction of total emissions across multiple basins in the united states. Proceedings of the National Academy of Scie...

    work page 2022

Show all 76 references
  1. [9]

    Learning roi transformer for oriented object detection in aerial images

    Jian Ding, Nan Xue, Yang Long, Gui-Song Xia, and Qikai Lu. Learning roi transformer for oriented object detection in aerial images. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 2849– 2858, 2019

    2019

  2. [10]

    Centernet: Keypoint triplets for object detection

    Kaiwen Duan, Song Bai, Lingxi Xie, Honggang Qi, Qing- ming Huang, and Qi Tian. Centernet: Keypoint triplets for object detection. In 2019 IEEE/CVF International Confer- ence on Computer Vision (ICCV), pages 6568–6577, 2019

    2019

  3. [11]

    Global tracking and quantification of oil and gas methane emissions from recurrent sentinel-2 imagery

    Thibaud Ehret, Aur ´elien De Truchis, Matthieu Mazzolini, Jean-Michel Morel, Alexandre D’aspremont, Thomas Lau- vaux, Riley Duren, Daniel Cusworth, and Gabriele Facciolo. Global tracking and quantification of oil and gas methane emissions from recurrent sentinel-2 imagery. Env...

    2022

  4. [12]

    A discriminatively trained, multiscale, deformable part model

    Pedro Felzenszwalb, David McAllester, and Deva Ra- manan. A discriminatively trained, multiscale, deformable part model. In 2008 IEEE Conference on Computer Vision and Pattern Recognition, pages 1–8, 2008

    2008

  5. [13]

    Felzenszwalb, Ross B

    Pedro F. Felzenszwalb, Ross B. Girshick, David McAllester, and Deva Ramanan. Object detection with discriminatively trained part-based models. IEEE transactions on pattern analysis and machine intelligence, 32(9):1627–1645, 2009

    2009

  6. [14]

    A decision-theoretic generalization of on-line learning and an application to boosting

    Yoav Freund and Robert E Schapire. A decision-theoretic generalization of on-line learning and an application to boosting. Journal of Computer and System Sciences , 55(1): 119–139, 1997

    1997

  7. [15]

    Rich feature hierarchies for accurate object detection and semantic segmentation

    Ross Girshick, Jeff Donahue, Trevor Darrell, and Jitendra Malik. Rich feature hierarchies for accurate object detection and semantic segmentation. In Proceedings of the IEEE con- ference on computer vision and pattern recognition , pages 580–587, 2014

    2014

  8. [16]

    Girshick, Forrest N

    Ross B. Girshick, Forrest N. Iandola, Trevor Darrell, and Jitendra Malik. Deformable part models are convolutional neural networks. CoRR, abs/1409.5403, 2014

    2014 arXiv

  9. [17]

    Redet: A rotation-equivariant detector for aerial object detection

    Jiaming Han, Jian Ding, Nan Xue, and Gui-Song Xia. Redet: A rotation-equivariant detector for aerial object detection. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 2786–2795, 2021

    2021

  10. [18]

    Multisized object detection using space- borne optical imagery

    Muhammad Haroon, Muhammad Shahzad, and Muham- mad Moazam Fraz. Multisized object detection using space- borne optical imagery. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 13:3032– 3046, 2020

    2020

  11. [19]

    Mask r-cnn

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

    2017

  12. [20]

    Mutdet: Mutually optimizing pre-training for remote sensing object detection, 2024

    Ziyue Huang, Yongchao Feng, Qingjie Liu, and Yunhong Wang. Mutdet: Mutually optimizing pre-training for remote sensing object detection, 2024

    2024

  13. [21]

    Methane and climate change – global methane tracker 2022 – analysis

    IEA. Methane and climate change – global methane tracker 2022 – analysis. https://www.iea.org/reports/ global - methane - tracker - 2022 / methane - and-climate-change, 2022

    2022

  14. [22]

    BD ORTHO®: L’image g ´eographique du ter- ritoire national, la France vue du ciel

    IGN - Institut national de l’information g ´eographique et foresti`ere. BD ORTHO®: L’image g ´eographique du ter- ritoire national, la France vue du ciel. https : / / geoservices.ign.fr/bdortho , 2024. Accessed: February 3, 2025

    2024

  15. [23]

    Acc `es `a l’Open Data Couvertures Spot 6-7 France

    Info Spot 6-7. Acc `es `a l’Open Data Couvertures Spot 6-7 France. https : / / www . geoportail . gouv . fr/actualites/observer- l- evolution- des- territoires-avec-spot , 2023. Accessed: June 10, 2025

    2023

  16. [24]

    Ultralytics YOLO, 2023

    Glenn Jocher, Jing Qiu, and Ayush Chaurasia. Ultralytics YOLO, 2023

    2023

  17. [25]

    Few-shot object detection via feature reweighting

    Bingyi Kang, Zhuang Liu, Xin Wang, Fisher Yu, Jiashi Feng, and Trevor Darrell. Few-shot object detection via feature reweighting. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 8420–8429, 2019

    2019

  18. [26]

    Yolov11: An overview of the key architectural enhancements, 2024

    Rahima Khanam and Muhammad Hussain. Yolov11: An overview of the key architectural enhancements, 2024

    2024

  19. [27]

    Cornernet: Detecting objects as paired keypoints

    Hei Law and Jia Deng. Cornernet: Detecting objects as paired keypoints. In Proceedings of the European confer- ence on computer vision, pages 734–750, 2018

    2018

  20. [28]

    Rotation- insensitive and context-augmented object detection in re- mote sensing images

    Ke Li, Gong Cheng, Shuhui Bu, and Xiong You. Rotation- insensitive and context-augmented object detection in re- mote sensing images. IEEE Transactions on Geoscience and Remote Sensing, 56(4):2337–2348, 2017. 9

    2017

  21. [29]

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

    Ke Li, Gang Wan, Gong Cheng, Liqiu Meng, and Junwei Han. Object detection in optical remote sensing images: A survey and a new benchmark. ISPRS Journal of Photogram- metry and Remote Sensing, 159:296–307, 2020

    2020

  22. [30]

    Large selective kernel network for remote sensing object detection, 2023

    Yuxuan Li, Qibin Hou, Zhaohui Zheng, Ming-Ming Cheng, Jian Yang, and Xiang Li. Large selective kernel network for remote sensing object detection, 2023

    2023

  23. [31]

    Lsknet: A foun- dation lightweight backbone for remote sensing, 2024

    Yuxuan Li, Xiang Li, Yimian Dai, Qibin Hou, Li Liu, Yongx- iang Liu, Ming-Ming Cheng, and Jian Yang. Lsknet: A foun- dation lightweight backbone for remote sensing, 2024

    2024

  24. [32]

    Belongie, Lubomir D

    Tsung-Yi Lin, Michael Maire, Serge J. Belongie, Lubomir D. Bourdev, Ross B. Girshick, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C. Lawrence Zitnick. Microsoft COCO: common objects in context. CoRR, abs/1405.0312, 2014

    2014 arXiv

  25. [33]

    The danish national effort to minimise methane emissions from biogas plants

    Anders Michael Fredenslund, Einar Gudmundsson, Julie Maria Falk, and Charlotte Scheutz. The danish national effort to minimise methane emissions from biogas plants. Waste Management, 157:321–329, 2023

    2023

  26. [34]

    Nathan Mundhenk, Goran Konjevod, Wesam A

    T. Nathan Mundhenk, Goran Konjevod, Wesam A. Sakla, and Kofi Boakye. A large contextual dataset for classifi- cation, detection and counting of cars with deep learning. In Computer Vision – ECCV 2016 , pages 785–800, Cham,

    2016

  27. [35]

    Jeffri Murrugarra-LLerena, Jose Henrique Lima Marques, and Claudio R. Jung. Gaucho: Gaussian distributions with cholesky decomposition for oriented object detection, 2025

    2025

  28. [36]

    Tech- nical Summary, page 35–144

    Intergovernmental Panel on Climate Change (IPCC). Tech- nical Summary, page 35–144. Cambridge University Press, 2023

    2023

  29. [37]

    Automatic methane plumes detection in time series of sentinel-5p l1b images

    Elyes Ouerghi, Thibaud Ehret, Carlo de Franchis, Gabriele Facciolo, Thomas Lauvaux, Enric Meinhardt, and Jean- Michel Morel. Automatic methane plumes detection in time series of sentinel-5p l1b images. ISPRS Annals of the Pho- togrammetry, Remote Sensing and Spatial Informatio...

    2022

  30. [38]

    Methane plumes detection on prisma l1 images with the adjusted spectral matched filter and wind data

    Elyes Ouerghi, Thibaud Ehret, Gabriele Facciolo, Enric Meinhardt, Jean-Michel Morel, Carlo De Franchis, and Thomas Lauvaux. Methane plumes detection on prisma l1 images with the adjusted spectral matched filter and wind data. In IGARSS 2023-2023 IEEE International Geoscience a...

    2023

  31. [39]

    Satellite observations reveal extreme methane leakage from a natural gas well blowout

    Sudhanshu Pandey, Ritesh Gautam, Sander Houweling, Hugo Denier Van Der Gon, Pankaj Sadavarte, Tobias Bors- dorff, Otto Hasekamp, Jochen Landgraf, Paul Tol, Tim Van Kempen, et al. Satellite observations reveal extreme methane leakage from a natural gas well blowout. Proceed- in...

    2019

  32. [40]

    Deep learning based electric pylon detection in remote sensing images

    Sijia Qiao, Yu Sun, and Haopeng Zhang. Deep learning based electric pylon detection in remote sensing images. Re- mote Sensing, 12(11):1857, 2020

    2020

  33. [41]

    Ng, and Robert B

    Neel Ramachandran, Jeremy Irvin, Mark Omara, Ritesh Gautam, Kelsey Meisenhelder, Erfan Rostami, Hao Sheng, Andrew Y . Ng, and Robert B. Jackson. Deep learning for detecting and characterizing oil and gas well pads in satellite imagery. Nature Communications, 15(1):7036, 2024

    2024

  34. [42]

    Vehicle detec- tion in aerial imagery : A small target detection benchmark

    Sebastien Razakarivony and Frederic Jurie. Vehicle detec- tion in aerial imagery : A small target detection benchmark. Journal of Visual Communication and Image Representation, 34:187–203, 2016

    2016

  35. [43]

    Yolov3: An incremental improvement

    Joseph Redmon and Ali Farhadi. Yolov3: An incremental improvement. CoRR, abs/1804.02767, 2018

    2018 arXiv

  36. [44]

    Girshick, and Ali Farhadi

    Joseph Redmon, Santosh Kumar Divvala, Ross B. Girshick, and Ali Farhadi. You only look once: Unified, real-time ob- ject detection. CoRR, abs/1506.02640, 2015

    2015 arXiv

  37. [45]

    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 process- ing systems, 28, 2015

    2015

  38. [46]

    Faster r-cnn: Towards real-time object detection with region proposal networks, 2016

    Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster r-cnn: Towards real-time object detection with region proposal networks, 2016

    2016

  39. [47]

    Anderson, Juan M

    Caleb Robinson, Ben Chugg, Brandon R. Anderson, Juan M. Lavista Ferres, and Daniel E. Ho. Mapping industrial poultry operations at scale with deep learning and aerial im- agery. CoRR, abs/2112.10988, 2021

    2021 arXiv

  40. [48]

    Lavista Ferres

    Caleb Robinson, Anthony Ortiz, Allen Kim, Rahul Dodhia, Andrew Zolli, Shivaprakash K Nagaraju, James Oakleaf, Joe Kiesecker, and Juan M. Lavista Ferres. Global renewables watch: A temporal dataset of solar and wind energy derived from satellite imagery. 2025

    2025

  41. [49]

    Schapire

    Robert E. Schapire. The boosting approach to machine learn- ing: An overview. Nonlinear estimation and classification , pages 149–171, 2003

    2003

  42. [50]

    Drew Shindell, Johan C. I. Kuylenstierna, Elisabetta Vig- nati, Rita van Dingenen, Markus Amann, Zbigniew Klimont, Susan C. Anenberg, Nicholas Muller, Greet Janssens- Maenhout, Frank Raes, Joel Schwartz, Greg Faluvegi, Luca Pozzoli, Kaarle Kupiainen, Lena H ¨oglund-Isaksson, L...

    2012

  43. [51]

    A har- monised, high-coverage, open dataset of solar photovoltaic installations in the uk

    Dan Stowell, Jack Kelly, Damien Tanner, Jamie Taylor, Ethan Jones, James Geddes, and Ed Chalstrey. A har- monised, high-coverage, open dataset of solar photovoltaic installations in the uk. Scientific Data, 7(1):394, 2020

    2020

  44. [52]

    Pbnet: Part-based convolutional neural network for complex composite object detection in remote sensing im- agery

    Xian Sun, Peijin Wang, Cheng Wang, Yingfei Liu, and Kun Fu. Pbnet: Part-based convolutional neural network for complex composite object detection in remote sensing im- agery. ISPRS Journal of Photogrammetry and Remote Sens- ing, 173:50–65, 2021

    2021

  45. [53]

    Fair1m: A benchmark dataset for fine- grained object recognition in high-resolution remote sens- ing imagery

    Xian Sun, Peijin Wang, Zhiyuan Yan, Feng Xu, Ruiping Wang, Wenhui Diao, Jin Chen, Jihao Li, Yingchao Feng, Tao Xu, et al. Fair1m: A benchmark dataset for fine- grained object recognition in high-resolution remote sens- ing imagery. ISPRS Journal of Photogrammetry and Remote Se...

    2022

  46. [54]

    A harmonised, high- coverage, open dataset of solar photovoltaic installations in the UK

    Scientific Data Curation Team. A harmonised, high- coverage, open dataset of solar photovoltaic installations in the UK. Nature, 2020. 10

    2020

  47. [55]

    Thorpe, Robert O

    Andrew K. Thorpe, Robert O. Green, David R. Thomp- son, Philip G. Brodrick, John W. Chapman, Clayton D. El- der, Itziar Irakulis-Loitxate, Daniel H. Cusworth, Alana K. Ayasse, Riley M. Duren, et al. Attribution of individual methane and carbon dioxide emission sources using em...

    2023

  48. [56]

    Fcos: Fully convolutional one-stage object detection

    Zhi Tian, Chunhua Shen, Hao Chen, and Tong He. Fcos: Fully convolutional one-stage object detection. In 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pages 9626–9635, 2019

    2019

  49. [57]

    Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors, 2022

    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, 2022

    2022

  50. [58]

    Yolov9: Learning what you want to learn using pro- grammable gradient information, 2024

    Chien-Yao Wang, I-Hau Yeh, and Hong-Yuan Mark Liao. Yolov9: Learning what you want to learn using pro- grammable gradient information, 2024

    2024

  51. [59]

    Williams, Mark Omara, Anthony Himmelberger, Daniel Zavala-Araiza, Katlyn MacKay, Joshua Benmergui, Maryann Sargent, Steven C

    James P. Williams, Mark Omara, Anthony Himmelberger, Daniel Zavala-Araiza, Katlyn MacKay, Joshua Benmergui, Maryann Sargent, Steven C. Wofsy, Steven P. Hamburg, and Ritesh Gautam. Small emission sources in aggregate dis- proportionately account for a large majority of total me...

    2025

  52. [60]

    Detectron2

    Yuxin Wu, Alexander Kirillov, Francisco Massa, Wan-Yen Lo, and Ross Girshick. Detectron2. https://github. com/facebookresearch/detectron2, 2019

    2019

  53. [61]

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

    Gui-Song Xia, Xiang Bai, Jian Ding, Zhen Zhu, Serge Be- longie, Jiebo Luo, Mihai Datcu, Marcello Pelillo, and Liang- pei Zhang. Dota: A large-scale dataset for object detection in aerial images. In Proceedings of the IEEE conference on computer vision and pattern recognition, ...

    2018

  54. [62]

    A review of object de- tection based on deep learning

    Youzi Xiao, Zhiqiang Tian, Jiachen Yu, Yinshu Zhang, Shuai Liu, Shaoyi Du, and Xuguang Lan. A review of object de- tection based on deep learning. Multimedia Tools and Appli- cations, 79:23729–23791, 2020

    2020

  55. [63]

    Oriented R-CNN for object detection

    Xingxing Xie, Gong Cheng, Jiabao Wang, Xiwen Yao, and Junwei Han. Oriented R-CNN for object detection. CoRR, abs/2108.05699, 2021

    2021 arXiv

  56. [64]

    Gliding vertex on the horizontal bounding box for multi-oriented object detec- tion

    Yongchao Xu, Mingtao Fu, Qimeng Wang, Yukang Wang, Kai Chen, Gui-Song Xia, and Xiang Bai. Gliding vertex on the horizontal bounding box for multi-oriented object detec- tion. IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(4):1452–1459, 2021

    2021

  57. [65]

    Arbitrary-oriented object de- tection with circular smooth label

    Xue Yang and Junchi Yan. Arbitrary-oriented object de- tection with circular smooth label. In Computer Vision– ECCV 2020: 16th European Conference, Glasgow, UK, Au- gust 23–28, 2020, Proceedings, Part VIII 16, pages 677–694. Springer, 2020

    2020

  58. [66]

    On the arbitrary-oriented ob- ject detection: Classification based approaches revisited

    Xue Yang and Junchi Yan. On the arbitrary-oriented ob- ject detection: Classification based approaches revisited. In- ternational Journal of Computer Vision, 130(5):1340–1365, 2022

    2022

  59. [67]

    R3det: Refined single-stage detector with feature refinement for ro- tating object

    Xue Yang, Junchi Yan, Ziming Feng, and Tao He. R3det: Refined single-stage detector with feature refinement for ro- tating object. In Proceedings of the AAAI conference on ar- tificial intelligence, pages 3163–3171, 2021

    2021

  60. [68]

    Rethinking rotated object detection with gaussian wasserstein distance loss

    Xue Yang, Junchi Yan, Qi Ming, Wentao Wang, Xiaopeng Zhang, and Qi Tian. Rethinking rotated object detection with gaussian wasserstein distance loss. In International con- ference on machine learning , pages 11830–11841. PMLR, 2021

    2021

  61. [69]

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

    Xue Yang, Xiaojiang Yang, Jirui Yang, Qi Ming, Wentao Wang, Qi Tian, and Junchi Yan. Learning high-precision bounding box for rotated object detection via kullback- leibler divergence. Advances in Neural Information Process- ing Systems, 34:18381–18394, 2021

    2021

  62. [70]

    Reppoints: Point set representation for object detection

    Ze Yang, Shaohui Liu, Han Hu, Liwei Wang, and Stephen Lin. Reppoints: Point set representation for object detection. In 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pages 9656–9665, 2019

    2019

  63. [71]

    Phase-shifting coder: Predicting ac- curate orientation in oriented object detection

    Yi Yu and Feipeng Da. Phase-shifting coder: Predicting ac- curate orientation in oriented object detection. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13354–13363, 2023

    2023

  64. [72]

    Girshick, and Trevor Darrell

    Ning Zhang, Jeff Donahue, Ross B. Girshick, and Trevor Darrell. Part-based r-cnns for fine-grained category detec- tion. CoRR, abs/1407.3867, 2014

    2014 arXiv

  65. [73]

    Object detection with deep learning: A review

    Zhong-Qiu Zhao, Peng Zheng, Shou-tao Xu, and Xindong Wu. Object detection with deep learning: A review. IEEE transactions on neural networks and learning systems , 30 (11):3212–3232, 2019

    2019

  66. [74]

    Mmrotate: A rotated object detection benchmark using pytorch

    Yue Zhou, Xue Yang, Gefan Zhang, Jiabao Wang, Yanyi Liu, Liping Hou, Xue Jiang, Xingzhao Liu, Junchi Yan, Chengqi Lyu, Wenwei Zhang, and Kai Chen. Mmrotate: A rotated object detection benchmark using pytorch. In Proceedings of the 30th ACM International Conference on Multimedia , 2022

    2022

  67. [75]

    Convolutional neu- ral network based automatic object detection on aerial im- ages

    Igor ˇSevo and Aleksej Avramovic. Convolutional neu- ral network based automatic object detection on aerial im- ages. IEEE Geoscience and Remote Sensing Letters, 13:1–5, 2016. 11 Towards Large Scale Geostatistical Methane Monitoring with Part-based Object Detection Supplementa...

    2016

  68. [2016]

    Springer International Publishing

Pith tools

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