Pith. sign in

REVIEW 3 major objections 6 minor 38 references

UASTHN: Uncertainty-Aware Deep Homography Estimation for UAV Satellite-Thermal Geo-localization

T0 review · 3 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read A UAV can tell when its thermal-to-satellite alignment is untrustworthy by checking whether random crops of the thermal image produce agreeing homography predictions.

desk verdict A real crop-consensus uncertainty mechanism with sound recovery math, but unreported rejection thresholds, a confounded fine-tuning setup, and DC=128 results that invert the core assumption. read the letter →

arxiv 2502.01035 v2 pith:HKXEL3XU submitted 2025-02-03 cs.RO cs.CV

Jiuhong Xiao , Giuseppe Loianno This is my paper · ORCID
classification cs.ROcs.CV
keywords uncertaintyestimationdeephomographythermalgeo-localizationtest-timeaugmentationensemblesUAVnavigationsatellite-thermalalignmentfailuredetection
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

This paper argues that a UAV can tell when its thermal-to-satellite image alignment is unreliable by making random crops of the thermal image at test time and checking whether the recovered alignments agree. If the four-corner displacements predicted from different crops disagree strongly, the system rejects the match instead of trusting it. The authors claim this crop-consensus uncertainty signal, combined with deep ensembles for model uncertainty, is the first uncertainty-estimation method for cross-domain localization. They report that with the STHN two-stage homography network it cuts mean center error to 7.25 m with a 97.5% success rate within a 512 m search radius, and that it detects failures caused by textureless thermal images, corrupted images, geometric noise, self-similar maps, outdated maps, and images that extend beyond the satellite map.

What carries the argument

The central object is the crop-consensus uncertainty score $U^{\mathrm{TTA}}_{RS\to RT}$: after generating $N_C-1$ cropped views of the thermal image, each with crop offset $o_c$, the homography network predicts displacements whose standard deviation across views, taken per corner, is the data-uncertainty measure. The recovery step matters because cropped views predict displacements relative to the crop; equations (1)-(5) compose the homographies via DLT to map each prediction back to the original thermal frame. The mechanism's premise is that all cropped views share the same underlying homography, so genuine agreement indicates confidence, while disagreement flags textureless, corrupted, self-similar, or out-of-range inputs. Deep ensembles supply a second channel, model uncertainty, merged with the crop score by max, min, or sum.

What would settle it

Take a thermal image whose true alignment error is known to be large and whose appearance is textureless or self-similar, such as a dark desert scene with repeated dunes, and run CropTTA. If it assigns a low uncertainty score while the center error exceeds the acceptance threshold, the consensus proxy has failed. A systematic search over the Boson-nighttime test set for low-scored, high-error samples would settle whether the claim generalizes; the paper's ROC analysis at DC=512 reports AUC=0.96, so the residual false-acceptance cases are the ones to inspect.

Watch

Extended reading notes

Core claim

The central discovery is that data uncertainty in deep homography estimation can be measured without retraining or auxiliary networks: crop the thermal image into several views with a known offset, run the existing homography network on each, transform each predicted displacement back to the original frame using a DLT-based composition of homographies, and take the standard deviation of the recovered displacements as the uncertainty score. A high score triggers rejection of the alignment. The paper shows this CropTTA mechanism outperforms or matches a direct modeling baseline and deep ensembles across three homography networks (DHN, IHN, and STHN), and that combining it with deep ensembles gives a fuller data-plus-model uncertainty picture. The reported headline result is a mean center error of 7.25 m and a success rate of 97.5% for STHN two-stage with CropTTA at a 512 m search radius, with the strongest gains at large search radii where alignment failures have a long-tailed error distribution.

Load-bearing premise

The load-bearing assumption is that the spread of predicted corner displacements across random crops tracks the true alignment error; if the network is consistently and confidently wrong on every crop, the uncertainty score stays low and the failure is not flagged, and this proxy was validated only on the Boson-nighttime dataset.

Editorial extensions

If this is right

  • A real-time UAV navigation system can run CropTTA at inference and reject low-confidence alignments instead of feeding drifted poses to the controller; with early stopping, STHN two-stage runs in about 78 ms on an RTX 2080 Ti.
  • At large search radii (DC=512 m), where failure errors are long-tailed, CropTTA gives the biggest accuracy and success-rate gains, suggesting uncertainty screening matters most for low-frequency localization.
  • Because CropTTA only wraps the existing homography network with crops, it can be added to any deep homography estimator without retraining an uncertainty module, though fine-tuning with a crop-augmented loss improves it.
  • Combining CropTTA with deep ensembles yields a comprehensive uncertainty estimate that captures both hard-to-align data and model disagreement, with max as the chosen merge function.

Reading between the lines

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

  • The same crop-consensus principle should transfer to other cross-modal registration tasks, such as RGB-thermal or radar-satellite alignment, because it never depends on the modality-specific appearance; a testable extension would be to run it on a daytime or seasonal variant of the same data.
  • The rejection threshold $s_c$ is a free parameter; a natural extension is to set it adaptively from a validation set to meet a target success rate, rather than fixing it, as the paper's own future-work section hints.
  • The uncertainty score could feed a sensor-fusion layer that downweights or re-plans when thermal geo-localization disagrees with other navigation cues, rather than only discarding the sample; this is an implicit consequence of treating the score as a reliability measure.
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

3 major / 6 minor

Summary. The paper proposes UASTHN, an uncertainty estimation framework for deep homography estimation (DHE) in UAV satellite-thermal geo-localization. The main component, CropTTA, measures data uncertainty as the standard deviation of recovered four-corner displacements across random crops of the thermal image at test time (Eq. (6)), and optionally combines this with deep ensembles (DE) for model uncertainty. A rejection threshold sc is used to discard high-uncertainty predictions, and success rates (SR) are reported as the proportion of non-rejected samples. Experiments on the Boson-nighttime dataset compare CropTTA against DE and direct modeling (DM) across three DHE backbones (DHN, IHN, STHN) at three search radii (DC = 128, 256, 512 m). The headline result is a mean center error of 7.25 m with a 97.5% success rate for STHN with CropTTA at DC = 512 m. The crop-recovery mathematics (Eqs. (1)-(5)) is sound, and the empirical comparison is extensive across backbones and uncertainty baselines. However, the rejection threshold sc is never stated, the DC = 128 m results appear to contradict the central assumption that crop-consistency standard deviation tracks alignment error, and the fine-tuning protocol confounds the effect of rejection with the effect of training with the crop loss.

Significance. If the claims hold, CropTTA would be a simple, backbone-agnostic, and efficient way to provide data-uncertainty estimates for thermal geo-localization, which is safety-relevant for UAV navigation in GPS-denied or low-light conditions. The idea of using crop-consistency as an uncertainty signal is a reasonable and novel contribution for this task, and the paper ships public code/models, includes extensive ablations, and evaluates with ROC curves, which is a standard and non-circular protocol. However, the current evidence is not fully convincing: the success-rate numbers are operating points on an unreported threshold curve, the DC = 128 m regime shows a direct contradiction with the assumed variance-error relationship, and the comparison to baselines is confounded by fine-tuning. These issues are load-bearing for the central claim that CropTTA provides effective and general data-uncertainty measurement, and they require additional analysis and reporting.

major comments (3)
  1. [Section III-B and Section V-B, Table I] The rejection threshold sc is defined in Section III-B but its value is never reported anywhere in the paper, including in the experiments or the supplementary material. Since every Success Rate in Table I and the headline "97.5% success rate" are defined as the proportion of samples not rejected by this threshold, the reported SR values are unconstrained operating points. Without sc, the reader cannot verify the rejection behavior, reproduce the results, or compare the operating points across methods. Please report the exact sc used for each DC and backbone, or alternatively present threshold-free metrics (e.g., ROC AUC with a justified failure definition, or a sweep over sc) as the primary evaluation.
  2. [Section V-B, Table I, Fig. 4] At DC = 128 m, CropTTA with STHN increases MACE from 7.51 m (no UE) to 8.26 m and CE from 6.66 m to 7.75 m while rejecting only 1.5% of samples (SR 98.5%). For the mean of the accepted samples to increase, the rejected samples must have had substantially below-average error, implying that the uncertainty score is inversely related to alignment error in this regime. This directly contradicts the assumption in Eq. (6) that crop-consistency standard deviation tracks alignment error. The paper's statement that "baselines without UE outperform, indicating inherent lower bounds" (Section V-B) does not explain this contradiction. Please provide an analysis of the DC = 128 m case, including the error distribution of rejected samples and a discussion of why rejection degrades accuracy, since this bears directly on the generality of CropTTA as a data-uncertainty estimator.
  3. [Section IV (Implementation Details) and Table I] The CropTTA models are first trained for 100k steps without CropTTA and then fine-tuned for an additional 200k steps with the crop loss LCropTTA (Eq. (8)), while the "no UE" baselines in Table I are presumably the unfine-tuned models. The comparison therefore conflates the effect of the rejection mechanism with the effect of fine-tuning with the crop loss. The paper never reports the fine-tuned model's no-rejection performance (i.e., using only the original displacement with no threshold applied). Please report MACE and CE for the fine-tuned model with no rejection, so that the incremental benefit of the rejection step can be isolated. Without this, the reported improvement at DC = 512 m cannot be attributed to the uncertainty estimation rather than to the fine-tuning protocol.
minor comments (6)
  1. [Abstract and Section V-B] The abstract states "geo-localization error of 7 m with a 97% success rate" while Table I reports CE = 7.25 m and SR = 97.5% for the same configuration; please make the numbers consistent or clarify the rounding.
  2. [Fig. 4 caption] The ROC analysis defines "expected rejected predictions" as predictions exceeding 25 m MACE, but this threshold is not justified. Please motivate this choice or show that the conclusions are robust to the threshold value.
  3. [Section III-C, Eq. (9)] Equation (9) defines U^DE_RS->RT, but the following sentence says "where U TTA ∈ R^{2×4} denotes the model uncertainty"; this should read U^DE.
  4. [Section III-C, Eq. (11) and Fig. 3f] The merge function f(·) is only described in Fig. 3f as choosing min, max, or add, and the default is stated to be max in the ablation text. Please state the default merge function explicitly in the main text near Eq. (11).
  5. [Section V-B, Table I] The number of ensemble models N_m used for the DE experiments is not stated, although it affects the DE uncertainty estimates and inference times; please report it in the implementation details.
  6. [Contributions, Section II] The claim that this is the first solution to address uncertainty estimation for localization using cross-domain data is strong; please provide a more precise comparison with prior uncertainty estimation methods in cross-domain localization to support the novelty claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the CropTTA uncertainty score is evaluated against held-out ground truth, and the prior STHN backbone/dataset are tools, not self-justifying premises.

full rationale

Walking the derivation chain, UASTHN's uncertainty estimate is not equivalent to its own inputs. The data-uncertainty score U_TTA (Eq. 6) is the standard deviation of four-corner displacements recovered from the original image plus N_C-1 cropped views (Eqs. 1-5); it is computed from the network's test-time forward passes and is never fitted to ground-truth errors, ground-truth MACE, or rejection outcomes. The claim that this score detects failures is evaluated on held-out test data (27K pairs, different regions and times) with an externally fixed 25 m MACE definition of a failure, reported as ROC AUC (Fig. 4: 0.96 at DC=512) and as accepted-sample MACE/CE in Table I, which is a non-circular protocol. The CropTTA loss (Eq. 8) is a training objective that regularizes the network toward crop-consistent displacements; the residual inconsistency at test time is a genuinely measured quantity, not a fitted parameter renamed as a prediction. The self-citations are not load-bearing: STHN [14] and the Boson dataset [13] are the authors' own prior tools used as backbone, baseline, and data, and the method is additionally demonstrated on the external DHN and IHN backbones, so the central claim does not reduce to a self-citation chain. The acknowledged DC=128 degradation (accepted-sample MACE rises from 7.51 m to 8.26 m after rejection) is an empirical failure of the variance-as-uncertainty assumption in the high-frequency regime, not a circular construction; the paper flags it openly as an 'inherent lower bound.' The unreported rejection threshold s_c and the missing fine-tuned no-rejection control are reproducibility/confound issues that belong to correctness risk, not to circularity. No step in the derivation equates a prediction with its input by construction.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central approach depends on the crop-consensus premise and the specific operating thresholds, both of which are empirical choices validated on one dataset. No new physical or mathematical entities are introduced.

free parameters (5)
  • crop offset o_c = 32 pixels (random sampling)
    Chosen from ablation (Fig. 3b); not derived, and affects the uncertainty distribution.
  • number of crops N_C = 5 at evaluation
    Selected from ablation (Fig. 3d); more than 5 gives diminishing returns.
  • rejection threshold s_c = not reported
    The rejection criterion in Section III-B; success rates in Table I and the headline 97% likely depend on this threshold, yet its value or selection method is not given.
  • MACE failure threshold for ROC = 25 m
    Used to define 'expected rejected predictions' in the ROC analysis (Section V-B, Fig. 4); chosen by the authors, not justified.
  • early stopping iteration count = first k iterations (value not specified in main text)
    Section III-C; reduces inference cost, but the exact k and its effect on uncertainty quality are under-specified.
assumptions (4)
  • domain assumption The satellite-thermal image pair is well modeled by a global planar homography, and any crop of the thermal image shares that same homography.
    Used in Section III-B, Eqs. (1)-(5), to recover displacements from cropped views; violated in non-planar scenes or with parallax.
  • ad hoc to paper Prediction variance across crops is a measure of data uncertainty.
    Core premise of CropTTA, stated as intuition in Section III-B; not derived from a probabilistic model or calibrated.
  • domain assumption The Boson-nighttime dataset is representative of thermal geo-localization conditions.
    All experiments in Section V use this single dataset; generalization to other cameras, times, and map sources is untested.
  • standard math Standard DLT and homography composition formulas are correct as used.
    Equations (1)-(5) rely on textbook DLT [36].

how reviews work

0 comments
Cite this review

Pith. "Pith review of UASTHN: Uncertainty-Aware Deep Homography Estimation for UAV Satellite-Thermal Geo-localization." pith.science (2026). https://pith.science/paper/HKXEL3XU

@misc{pith2026250201035,
  author       = {Pith},
  title        = {Pith review of: UASTHN: Uncertainty-Aware Deep Homography Estimation for UAV Satellite-Thermal Geo-localization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HKXEL3XU}},
  note         = {Machine review of arXiv:2502.01035}
}
read the original abstract

Geo-localization is an essential component of Unmanned Aerial Vehicle (UAV) navigation systems to ensure precise absolute self-localization in outdoor environments. To address the challenges of GPS signal interruptions or low illumination, Thermal Geo-localization (TG) employs aerial thermal imagery to align with reference satellite maps to accurately determine the UAV's location. However, existing TG methods lack uncertainty measurement in their outputs, compromising system robustness in the presence of textureless or corrupted thermal images, self-similar or outdated satellite maps, geometric noises, or thermal images exceeding satellite maps. To overcome these limitations, this paper presents UASTHN, a novel approach for Uncertainty Estimation (UE) in Deep Homography Estimation (DHE) tasks for TG applications. Specifically, we introduce a novel Crop-based Test-Time Augmentation (CropTTA) strategy, which leverages the homography consensus of cropped image views to effectively measure data uncertainty. This approach is complemented by Deep Ensembles (DE) employed for model uncertainty, offering comparable performance with improved efficiency and seamless integration with any DHE model. Extensive experiments across multiple DHE models demonstrate the effectiveness and efficiency of CropTTA in TG applications. Analysis of detected failure cases underscores the improved reliability of CropTTA under challenging conditions. Finally, we demonstrate the capability of combining CropTTA and DE for a comprehensive assessment of both data and model uncertainty. Our research provides profound insights into the broader intersection of localization and uncertainty estimation. The code and models are publicly available.

Figures

Figures reproduced from arXiv: 2502.01035 by the authors.

Figure 1
Figure 1. Data Uncertainty in Thermal Geo-localization (TG): Our approach captures six categories of high data-uncertainty [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. UASTHN framework: CropTTA augments thermal images, and network [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Ablation Study. We use success rate and Validation (Val) MACE metrics to ablate the training and evaluation settings. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: ROC curves for CropTTA with STHN two-stage [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: MACE histogram for CropTTA with STHN two-stage [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: CropTTA detected failure cases with the STHN two-stage method. Thermal images overlap with satellite images, [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 30 canonical work pages

  1. [1]

    A review on uav-based applications for precision agriculture,

    D. C. Tsouros, S. Bibi, and P. G. Sarigiannidis, “A review on uav-based applications for precision agriculture,” Information, vol. 10, no. 11, 2019

    work page 2019

  2. [2]

    Thermal and visual tracking of photovoltaic plants for autonomous uav inspection,

    L. Morando, C. T. Recchiuto, J. Calla, P. Scuteri, and A. Sgorbissa, “Thermal and visual tracking of photovoltaic plants for autonomous uav inspection,” Drones, vol. 6, no. 11, 2022

    work page 2022

  3. [3]

    Uav- assisted wireless localization for search and rescue,

    M. Atif, R. Ahmad, W. Ahmad, L. Zhao, and J. J. Rodrigues, “Uav- assisted wireless localization for search and rescue,” IEEE Systems Journal, vol. 15, no. 3, pp. 3261–3272, 2021

    work page 2021

  4. [4]

    QuadFormer: Quadruple Transformer for Unsupervised Domain Adaptation in Power Line Segmentation of Aerial Images

    P. P. Rao, F. Qiao, W. Zhang, Y . Xu, Y . Deng, G. Wu, and Q. Zhang, “Quadformer: Quadruple transformer for unsupervised domain adap- tation in power line segmentation of aerial images,” arXiv preprint arXiv:2211.16988, 2022

    work page Pith review arXiv 2022

  5. [5]

    Au- tonomous power line inspection with drones via perception-aware mpc,

    J. Xing, G. Cioffi, J. Hidalgo-Carri ´o, and D. Scaramuzza, “Au- tonomous power line inspection with drones via perception-aware mpc,” in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023, pp. 1086–1093

    work page 2023

  6. [6]

    Unifying foundation models with quadrotor control for visual track- ing beyond object categories,

    A. Saviolo, P. Rao, V . Radhakrishnan, J. Xiao, and G. Loianno, “Unifying foundation models with quadrotor control for visual track- ing beyond object categories,” in IEEE International Conference on Robotics and Automation (ICRA) , 2024, pp. 7389–7396

    work page 2024

  7. [7]

    A review on absolute visual localization for uav,

    A. Couturier and M. A. Akhloufi, “A review on absolute visual localization for uav,” Robotics and Autonomous Systems , vol. 135, p. 103666, 2021

    work page 2021

  8. [8]

    Foundloc: Vision-based onboard aerial localization in the wild,

    Y . He, I. Cisneros, N. Keetha, J. Patrikar, Z. Ye, I. Higgins, Y . Hu, P. Kapoor, and S. Scherer, “Foundloc: Vision-based onboard aerial localization in the wild,” arXiv preprint arXiv:2310.16299 , 2023

    arXiv 2023

Show all 38 references
  1. [9]

    A seasonally invariant deep transform for visual terrain-relative navigation,

    A. T. Fragoso, C. T. Lee, A. S. McCoy, and S.-J. Chung, “A seasonally invariant deep transform for visual terrain-relative navigation,” Science Robotics, vol. 6, no. 55, p. eabf3320, 2021

    2021

  2. [10]

    Visual localization with google earth images for robust global pose estimation of uavs,

    B. Patel, T. D. Barfoot, and A. P. Schoellig, “Visual localization with google earth images for robust global pose estimation of uavs,” in IEEE International Conference on Robotics and Automation (ICRA) , 2020, pp. 6491–6497

    2020

  3. [11]

    Google map aided visual navigation for uavs in gps-denied environ- ment,

    M. Shan, F. Wang, F. Lin, Z. Gao, Y . Z. Tang, and B. M. Chen, “Google map aided visual navigation for uavs in gps-denied environ- ment,” in IEEE International Conference on Robotics and Biomimetics (ROBIO), 2015, pp. 114–119

    2015

  4. [12]

    Caltech aerial rgb-thermal dataset in the wild,

    C. Lee, M. Anderson, N. Raganathan, X. Zuo, K. Do, G. Gkioxari, and S.-J. Chung, “Caltech aerial rgb-thermal dataset in the wild,”arXiv preprint arXiv:2403.08997, 2024

    2024 arXiv

  5. [13]

    Long-range uav thermal geo-localization with satellite imagery,

    J. Xiao, D. Tortei, E. Roura, and G. Loianno, “Long-range uav thermal geo-localization with satellite imagery,” in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023, pp. 5820– 5827

    2023

  6. [14]

    Sthn: Deep homogra- phy estimation for uav thermal geo-localization with satellite imagery,

    J. Xiao, N. Zhang, D. Tortei, and G. Loianno, “Sthn: Deep homogra- phy estimation for uav thermal geo-localization with satellite imagery,” IEEE Robotics and Automation Letters, vol. 9, no. 10, pp. 8754–8761, 2024

    2024

  7. [15]

    Deep learning,

    Y . LeCun, Y . Bengio, and G. Hinton, “Deep learning,” nature, vol. 521, no. 7553, pp. 436–444, 2015

    2015

  8. [16]

    Deep image homog- raphy estimation,

    D. DeTone, T. Malisiewicz, and A. Rabinovich, “Deep image homog- raphy estimation,” arXiv preprint arXiv:1606.03798 , 2016

    2016 arXiv

  9. [17]

    Iterative deep homog- raphy estimation,

    S.-Y . Cao, J. Hu, Z. Sheng, and H.-L. Shen, “Iterative deep homog- raphy estimation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2022, pp. 1879– 1888

    2022

  10. [18]

    Unsupervised deep homography: A fast and robust homography estimation model,

    T. Nguyen, S. W. Chen, S. S. Shivakumar, C. J. Taylor, and V . Kumar, “Unsupervised deep homography: A fast and robust homography estimation model,” IEEE Robotics and Automation Letters , vol. 3, no. 3, pp. 2346–2353, 2018

    2018

  11. [19]

    Localtrans: A multiscale local transformer network for cross-resolution homography estimation,

    R. Shao, G. Wu, Y . Zhou, Y . Fu, L. Fang, and Y . Liu, “Localtrans: A multiscale local transformer network for cross-resolution homography estimation,” in Proceedings of the IEEE/CVF international conference on computer vision (CVPR) , 2021, pp. 14 890–14 899

    2021

  12. [20]

    Multipoint: Cross-spectral registration of thermal and optical aerial imagery,

    F. Achermann, A. Kolobov, D. Dey, T. Hinzmann, J. J. Chung, R. Siegwart, and N. Lawrance, “Multipoint: Cross-spectral registration of thermal and optical aerial imagery,” in Proceedings of the 2020 Conference on Robot Learning, ser. Proceedings of Machine Learning Research, J....

    2020

  13. [21]

    Infrared and visible image ho- mography estimation using multiscale generative adversarial network,

    Y . Luo, X. Wang, Y . Wu, and C. Shu, “Infrared and visible image ho- mography estimation using multiscale generative adversarial network,” Electronics, vol. 12, no. 4, 2023

    2023

  14. [22]

    Coarse- to-fine homography estimation for infrared and visible images,

    X. Wang, Y . Luo, Q. Fu, Y . He, C. Shu, Y . Wu, and Y . Liao, “Coarse- to-fine homography estimation for infrared and visible images,” Elec- tronics, vol. 12, no. 21, 2023

    2023

  15. [23]

    A survey of uncertainty in deep neural networks,

    J. Gawlikowski, C. R. N. Tassi, M. Ali, J. Lee, M. Humt, J. Feng, A. Kruspe, R. Triebel, P. Jung, R. Roscher et al. , “A survey of uncertainty in deep neural networks,” Artificial Intelligence Review , vol. 56, no. Suppl 1, pp. 1513–1589, 2023

    2023

  16. [24]

    What uncertainties do we need in bayesian deep learning for computer vision?

    A. Kendall and Y . Gal, “What uncertainties do we need in bayesian deep learning for computer vision?” in Advances in Neural Informa- tion Processing Systems , vol. 30, 2017

    2017

  17. [25]

    A review of uncertainty quan- tification in deep learning: Techniques, applications and challenges,

    M. Abdar, F. Pourpanah, S. Hussain, D. Rezazadegan, L. Liu, M. Ghavamzadeh, P. Fieguth, X. Cao, A. Khosravi, U. R. Acharya, V . Makarenkov, and S. Nahavandi, “A review of uncertainty quan- tification in deep learning: Techniques, applications and challenges,” Information Fusio...

    2021

  18. [26]

    Bet- ter aggregation in test-time augmentation,

    D. Shanmugam, D. Blalock, G. Balakrishnan, and J. Guttag, “Bet- ter aggregation in test-time augmentation,” in Proceedings of the IEEE/CVF international conference on computer vision , 2021, pp. 1214–1223

    2021

  19. [27]

    Understanding test-time augmentation,

    M. Kimura, “Understanding test-time augmentation,” in International Conference on Neural Information Processing . Springer, 2021, pp. 558–569

    2021

  20. [28]

    Learning loss for test-time augmenta- tion,

    I. Kim, Y . Kim, and S. Kim, “Learning loss for test-time augmenta- tion,” in Advances in Neural Information Processing Systems , vol. 33, 2020, pp. 4163–4174

    2020

  21. [29]

    Memo: Test time robustness via adaptation and augmentation,

    M. Zhang, S. Levine, and C. Finn, “Memo: Test time robustness via adaptation and augmentation,” in Advances in Neural Information Processing Systems, vol. 35, 2022, pp. 38 629–38 642

    2022

  22. [30]

    Simple and scalable predictive uncertainty estimation using deep ensembles,

    B. Lakshminarayanan, A. Pritzel, and C. Blundell, “Simple and scalable predictive uncertainty estimation using deep ensembles,” in Advances in neural information processing systems , vol. 30, 2017

    2017

  23. [31]

    Uncertainty quantification and deep ensembles,

    R. Rahaman and A. Thiery, “Uncertainty quantification and deep ensembles,” in Advances in Neural Information Processing Systems , M. Ranzato, A. Beygelzimer, Y . Dauphin, P. Liang, and J. W. Vaughan, Eds., vol. 34. Curran Associates, Inc., 2021, pp. 20 063–20 075

    2021

  24. [32]

    Deep ensembles: A loss landscape perspective,

    S. Fort, H. Hu, and B. Lakshminarayanan, “Deep ensembles: A loss landscape perspective,” arXiv preprint arXiv:1912.02757 , 2019

    1912 arXiv

  25. [33]

    Deep ensembles work, but are they necessary?

    T. Abe, E. K. Buchanan, G. Pleiss, R. Zemel, and J. P. Cunningham, “Deep ensembles work, but are they necessary?” in Advances in Neural Information Processing Systems , vol. 35, 2022, pp. 33 646– 33 660

    2022

  26. [34]

    Hvc-net: Unifying homography, visibility, and confidence learning for planar object tracking,

    H. Zhang and Y . Ling, “Hvc-net: Unifying homography, visibility, and confidence learning for planar object tracking,” in European Conference on Computer Vision . Springer, 2022, pp. 701–718

    2022

  27. [35]

    Cuahn-vio: Content-and-uncertainty-aware homography network for visual-inertial odometry,

    Y . Xu and G. C. de Croon, “Cuahn-vio: Content-and-uncertainty-aware homography network for visual-inertial odometry,” arXiv preprint arXiv:2208.13935, 2022

    2022 arXiv

  28. [36]

    Direct linear transformation from comparator coordinates into object space coordinates in close- range photogrammetry,

    Y . Abdel-Aziz, H. Karara, and M. Hauck, “Direct linear transformation from comparator coordinates into object space coordinates in close- range photogrammetry,” Photogrammetric Engineering & Remote Sensing, vol. 81, no. 2, pp. 103–107, 2015

    2015

  29. [37]

    A review and comparative study on probabilistic object detection in autonomous driving,

    D. Feng, A. Harakeh, S. L. Waslander, and K. Dietmayer, “A review and comparative study on probabilistic object detection in autonomous driving,” IEEE Transactions on Intelligent Transportation Systems , vol. 23, no. 8, pp. 9961–9980, 2021

    2021

  30. [38]

    Decoupled weight decay regularization,

    I. Loshchilov and F. Hutter, “Decoupled weight decay regularization,” in International Conference on Learning Representations , 2019

    2019

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.