Pith. sign in

REVIEW 4 major objections 6 minor 55 references

Improving Out-of-Distribution Detection via Dynamic Covariance Calibration

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper proposes that distance-based OOD detection can be improved per input by shrinking the within-class covariance along the residual-space direction of each real-time feature, without retraining the model.

desk verdict A genuinely new test-time covariance adjustment for OOD scores with broad experiments, but the well-definedness proof does not cover the min over classes and the SOTA claim exceeds the table. read the letter →

arxiv 2506.09399 v3 pith:Z7V2KIJK submitted 2025-06-11 cs.CV

Kaiyu Guo , Zijian Wang , Tan Pan , Brian C. Lovell , Mahsa Baktashmotlagh This is my paper · ORCID
classification cs.CV
keywords out-of-distributiondetectionMahalanobisdistancedynamiccovariancecalibrationresidualspaceprojectionpost-hocOODinformationgeometryself-supervisedmodelsshrinkage
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 claims that the geometry used by distance-based out-of-distribution (OOD) detection can be corrected at test time instead of being taken as a static object from training data. Its proposal is to watch each new input feature and shrink the within-class covariance along that feature's projection onto the residual space of the training distribution, making the Mahalanobis-like distance more sensitive to deviations in exactly that direction. The paper argues that training outlier features distort the covariance along directions that new OOD features also occupy, so a dynamic, per-input adjustment separates in- and out-of-distribution samples better than a static one. If true, OOD detection for any pre-trained feature extractor improves in a purely post-hoc manner, with no retraining of the model. The claim is supported by experiments across CIFAR and ImageNet-1k backbones, including the self-supervised DINO model.

What carries the argument

The central object is the dynamically adjusted precision matrix $M(f) = (\Sigma_R - B^\top a a^\top B)^{-1}$, an online rank-one downdate of the within-class covariance before inversion. $B$ collects the residual-space eigenvectors of $\Sigma_R$, and $a = B^\top f$ projects the live feature onto that space. The Sherman-Morrison formula supplies the theoretical link: it shows how subtracting the outer product changes the inverse, and Theorem 4.2 converts the requirement that the adjusted form stay non-negative into the checkable inequalities $p>1$, $q>1$, $s\ge 0$. The scoring function then takes the smallest class-mean distance under this per-input geometry.

What would settle it

A controlled experiment on synthetic features would settle the central claim: train a Gaussian in-distribution model with an injected outlier cluster in direction $u$, then evaluate on OOD samples drawn from a direction $v$ orthogonal to $u$. If the dynamic score still beats static Mahalanobis, the outliers-align-with-novel-directions hypothesis is not the mechanism; if the advantage vanishes, the hypothesis is confirmed. Additionally, one could search for real feature spaces where some inputs give $p\le 1$ or $s<0$, which would make the adjusted matrix non-positive-definite and produce invalid distance scores.

Watch

Extended reading notes

Core claim

On its own terms, the paper's discovery is a score function that recalibrates the Mahalanobis geometry per test point. Starting from the within-class covariance matrix $\Sigma_R$ of centered training features, the score for a real-time feature $f$ is $s(f) = -\min_i \sqrt{r_i^\top(\Sigma_R - B^\top a a^\top B)^{-1} r_i}$, where $r_i = f - \mu_i$ is the deviation from class mean $i$, $B$ is the basis of the residual space (the eigenvectors of $\Sigma_R$ belonging to its smallest eigenvalues), and $a = B^\top f$ is the coefficient vector of $f$ in that residual space. Subtracting $B^\top a a^\top B$ removes the component of variance along $f$'s residual direction before the matrix is inverted, so the distance metric becomes locally more sensitive to novel directions while leaving the principal space untouched. A theorem gives conditions ($p>1$, $q>1$, $s\ge 0$) under which the adjusted quadratic form is a valid, non-negative distance; the paper reports that these conditions hold on its ImageNet experiments. The paper reports that this score improves both AUROC and FPR95 over the compared post-hoc detectors on seven pre-trained models.

Load-bearing premise

The method assumes that the directions in which training features are distorted by outliers are the same directions in which new OOD test features lie, and that for every input the quantities $p>1$, $q>1$, $s\ge 0$ hold so the adjusted covariance stays positive definite.

Editorial extensions

If this is right

  • Post-hoc OOD detection can be improved on a frozen pre-trained feature extractor by adding a per-input covariance update, so the method composes with existing backbones without any training phase.
  • The residual-space restriction preserves the principal structure of the in-distribution data, which is why the score stays meaningful on weaker encoders such as ResNet-50 where static subspace methods struggle.
  • The dynamic adjustment applies not only to Mahalanobis distance but to any covariance-induced distance; the appendix shows it also helps Euclidean distance and RMDS in near-OOD settings.
  • Because the condition for a valid distance holds empirically when class clusters are compact and far from the origin, the method is expected to transfer to other well-separated feature spaces, including self-supervised ones like DINO.

Reading between the lines

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

  • Editorial inference: the update is essentially a rank-one downdate of the precision matrix along the live sample's residual direction; one natural extension the paper does not develop is to accumulate several test-point downdates into a batch-level update, which would drift the geometry over a test stream.
  • Editorial inference: the central hypothesis that training outliers align with OOD directions could be probed directly by injecting synthetic outlier directions into Gaussian features; if OOD samples are drawn from a direction orthogonal to the injected outlier, the score's advantage over static Mahalanobis should shrink or vanish, and if it does not, the paper's explanation would need revision.
  • Editorial inference: the same dynamic-geometry idea should transfer to retrieval or open-set recognition tasks where each query defines a local direction along which the feature metric should be contracted.
  • Editorial inference: the OpenOOD near-OOD table shows the dynamic update helps RMDS but hurts plain Mahalanobis on some splits, so the practical claim is not that every distance improves, but that the geometry choice and the dynamic update interact; a pairing rule for which base distance to use is left open.
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 / 6 minor

Summary. The paper proposes Dynamic Covariance Calibration (DCC), a post-hoc out-of-distribution detection score that dynamically adjusts the within-class covariance matrix in the residual space of the training feature distribution using the real-time input feature. The score is s(f) = -min_i sqrt(r_i^T (Sigma_R - B^T a a^T B)^{-1} r_i), with r_i = f - mu_i, a the residual-space coefficient vector of f, B the residual basis, and Sigma_R the within-class covariance. The authors justify the construction with a matrix-perturbation theorem (Theorem 4.2) and evaluate on CIFAR (DenseNet, WideResNet) and ImageNet-1k (ViT, ResNet-50, Swin-B, DeiT, DINO) benchmarks, reporting strong AUROC and FPR95 numbers and claiming state-of-the-art performance.

Significance. The core idea is simple, test-time only, and computationally cheap, and the empirical study is broad, covering multiple architectures including a self-supervised DINO model, with code released. These are genuine strengths. However, the theoretical guarantee as stated does not cover the actual algorithm, because Theorem 4.2 certifies only a single anchor vector while the score takes a minimum over all class residuals; for p>1 the working matrix is indefinite and the score may be undefined. The reported SOTA claim is also overstated for the ViT row in Table 2, and the residual dimension is selected from test performance in Section 5.6. If the well-definedness issue is resolved and the benchmarking is made more rigorous (error bars, fixed validation protocol), the method would be a useful contribution to post-hoc OOD detection.

major comments (4)
  1. [§4.1, §4.3, Algorithm 1] Theorem 4.2 only proves d(f) >= 0 for a single anchor a, but the score in Eq. (3) requires r_i^T M(f) r_i >= 0 for every class residual r_i = f - mu_i before taking the minimum. For p > 1, the Sherman-Morrison determinant shows Sigma_R - B^T a a^T B is indefinite, so some quadratic forms may be negative and the square root in Algorithm 1 may be undefined. Appendix B.1 (Figure 8) verifies s >= 0 only for the closest class mean, not for all N_c classes, and no figure verifies the actual inequality (s-1)^2 <= (p-1)(q-1). The paper must either prove the condition for every class residual or modify the algorithm (e.g., clipping or explicit positive-definiteness check) so the reported score is always well-defined.
  2. [§5.6, Tables 1-2] The residual space dimension k is selected from the test AUROC: Figures 4 and 5 show performance varying with dimensionality, and the main tables report numbers at the best-performing dimension. No error bars, confidence intervals, or multiple-seed runs are reported in the main tables. This makes the comparison with fixed baselines optimistic and prevents assessing whether the reported improvements are statistically significant. Please fix the dimension by a proper validation protocol or report sensitivity and variance.
  3. [§5.3, Table 2] The claim that the method 'achieves state-of-the-art performance on both AUROC and FPR95 across all four pre-trained models' is contradicted by Table 2 itself: on ViT, WDiscOOD has AUROC 94.41 vs ours 94.27 and FPR95 26.35 vs ours 26.94. In addition, the average row excludes WDiscOOD for Swin-B because of NaN scores, so the comparison is not on identical model sets. The SOTA claim should be qualified accordingly.
  4. [§4.3, Appendix B.1, §5.7] The validity condition of Theorem 4.2 is only empirically motivated, not verified for the score actually computed. Section 4.3 asserts p >> 1 and q >> 1 and that s >= 0, and Section 5.7 shows distributions of p and q, but the condition of the theorem is the joint inequality (s-1)^2 <= (p-1)(q-1); the s values in Appendix B.1 are shown only for the closest class and are not combined with p and q to check the inequality. Since the positive-definiteness of M(f) is load-bearing for the method, this gap should be addressed directly, either by a sharper proof for the min over classes or by a per-class empirical check reported in the main text.
minor comments (6)
  1. [Appendix A] The heading 'Theorical Proof' should read 'Theoretical Proof'.
  2. [Tables 8-9] The column header 'AVG' should be 'Avg.', and 'Comparision' is misspelled in the captions of Tables 1 and 8.
  3. [§4.2] The text alternates between 'principle space' and 'principal space'; the standard term is 'principal space'.
  4. [§5.5] The sentence 'the method without considering the real-time features can not perform better than the full method in all scenarios' is ambiguous and should be rephrased; 'can not' should be 'cannot' and the intended meaning clarified.
  5. [Appendix F] The statement doubting the validity of WiscOOD's CLIP experiments, based on an unverified possible overlap between the image-text training set and OOD datasets, is speculative and should either be supported with evidence or removed.
  6. [Figure 8, Appendix B.1] The caption 'The s values on ImageNet-1k pre-trained ResNet-50 and ViT' describes distributions rather than values; also, the x-axis in panel (b) starts at 0, which makes it difficult to confirm that all values are strictly positive.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the dynamic covariance score is computed from pre-trained statistics and the test feature, with no parameter fitted to OOD labels and no imported uniqueness theorem; the central claim does not reduce to the one overlapping-author citation.

full rationale

The central score s(f) in Eq. 3 is defined directly from pre-trained model statistics (within-class covariance Σ_R, class means μ_i, residual basis B) and the real-time feature f, with no parameter fitted to OOD labels and no test-set information used in constructing the score. Theorem 4.2 is a standard Sherman-Morrison computation and is proved in Appendix A; it is not imported from prior work. The p>1 regime is a well-definedness and correctness concern rather than circularity: the paper states the condition (s-1)^2 ≤ (p-1)(q-1) explicitly and checks p, q, and s empirically in Figures 6 and 8. The only overlapping-author citation is Pan et al. (2023) in Section 4.2, used to motivate why residual-space projection avoids shared model distributions; that premise is also visualized in Figure 3 with ViT and DeiT features, and the residual-space construction itself follows the external VIM work (Wang et al., 2022). The central OOD-detection claim therefore stands on the paper's own equations and experiments, not on a self-citation chain. No specific circular reduction, such as an equation equaling its input by construction or a fitted parameter renamed as a prediction, can be exhibited from the paper's text.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The method's central claim rests on three domain assumptions: Gaussian per-class features, OOD energy concentrated in residual space, and the empirical positivity condition of Theorem 4.2. The only tuned numeric parameter is the residual dimension k, selected by test performance. No new entities are postulated.

free parameters (1)
  • residual space dimension k = not reported precisely; around 250 for CIFAR DenseNet in Figure 4
    The residual basis B keeps the k eigenvectors of the within-class covariance with smallest eigenvalues. Section 5.6 sweeps k and reports that the best AUROC occurs around 250 dimensions, implying k is selected by test performance rather than by a validation rule. The main tables do not state the k used for each backbone.
assumptions (4)
  • domain assumption Per-class feature distributions are approximately Gaussian, so within-class covariance defines the relevant geometry.
    Invoked in Section 4.1 to treat covariance as the geometric prior and in Section 4.3 to build the score from class means and within-class covariance.
  • domain assumption OOD features have most of their energy in the residual space of the ID covariance.
    Adopted from VIM and Neco in Section 4.2; the method confines all covariance adjustments to this space, so the claim inherits the assumption.
  • domain assumption For every test feature the conditions of Theorem 4.2 hold: p>1, q>1 and s>=0, making M(f) positive definite.
    The theorem only gives a conditional guarantee; the paper verifies the distributions of p and q on ImageNet backbones (Section 5.7) and s in Appendix B.1, but does not prove them for all inputs.
  • standard math Sherman-Morrison formula and eigendecomposition of symmetric matrices are valid.
    Used in the proof of Theorem 4.2 and in Algorithm 1.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improving Out-of-Distribution Detection via Dynamic Covariance Calibration." pith.science (2026). https://pith.science/paper/Z7V2KIJK

@misc{pith2026250609399,
  author       = {Pith},
  title        = {Pith review of: Improving Out-of-Distribution Detection via Dynamic Covariance Calibration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Z7V2KIJK}},
  note         = {Machine review of arXiv:2506.09399}
}
read the original abstract

Out-of-Distribution (OOD) detection is essential for the trustworthiness of AI systems. Methods using prior information (i.e., subspace-based methods) have shown effective performance by extracting information geometry to detect OOD data with a more appropriate distance metric. However, these methods fail to address the geometry distorted by ill-distributed samples, due to the limitation of statically extracting information geometry from the training distribution. In this paper, we argue that the influence of ill-distributed samples can be corrected by dynamically adjusting the prior geometry in response to new data. Based on this insight, we propose a novel approach that dynamically updates the prior covariance matrix using real-time input features, refining its information. Specifically, we reduce the covariance along the direction of real-time input features and constrain adjustments to the residual space, thus preserving essential data characteristics and avoiding effects on unintended directions in the principal space. We evaluate our method on two pre-trained models for the CIFAR dataset and five pre-trained models for ImageNet-1k, including the self-supervised DINO model. Extensive experiments demonstrate that our approach significantly enhances OOD detection across various models. The code is released at https://github.com/workerbcd/ooddcc.

Figures

Figures reproduced from arXiv: 2506.09399 by the authors.

Figure 1
Figure 1. Visualization of Mahalanobis distance score distributions with (w) and without (w/o) Dynamic Adjustment. We compare scores from the CIFAR-100 dataset (ID) and the SVHN dataset (OOD) using a CIFAR-100 pre-trained DenseNet. Without dy￾namic adjustment, the scores have high variance and a large over￾lapping region. With dynamic adjustment, the variance is lower and the overlap is reduced. ever, deep models are often we… view at source ↗
Figure 2
Figure 2. Visualization of the difference among distances calculated with no prior matrix, a static prior matrix, and a dynamic prior matrix. The distance with no prior matrix is Euclidean distance where M is an identity matrix. In the case of the static prior matrix, the geometry can not be adjusted, even if it is distorted by outlier features in the training samples. With the dynamic prior matrix, the geometry can be refine… view at source ↗
Figure 3
Figure 3. Visualization of the model distribution between ImageNet-1k pre-trained ViT and DeiT. We labeled the features from ViT with orange colour and DeiT with blue colour. The features from different models are distributed separately, which suggests the model distribution is also the principal information affecting the feature distribution. In this context, the feature from the unknown distribution in the penultimate layer… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: AUROC on CIFAR-10 and CIFAR-100 pre-trained DenseNet with different residual space dimension 5.6. Performance w.r.t Residual Dimensionality [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: shows the robustness of residual space in large￾scale OOD detection. The trend in the ViT experiments is consistent with the findings in VIM (Wang et al., 2022). 250 500 750 1000 1250 1500 1750 2000 dims 70.0 75.0 80.0 85.0 90.0 95.0 100.0 AUROC(%) Textures SUN Place i…
Figure 7
Figure 7. Figure 7: The l2 norms of features extracted by ImageNet-1k pre￾trained ResNet-50 and ViT. 6. Conclusion In this paper, we introduce a new perspective for defining distance-based OOD scores that dynamically refine the dis￾torted geometry of the training distribution. Building on…
Figure 6
Figure 6. Figure 6: p and q values on ImageNet-1k pre-trained ResNet-50 5.8. Vector Norms in Residual Space In addition to examining the effectiveness of projection in residual space based on performance metrics, we investigate further how residual projection helps OOD detection by analyz…
Figure 8
Figure 8. Figure 8: The s values on ImageNet-1k pre-trained ResNet-50 and ViT. B.2. Dynamic Adjustment in Euclidean Distance Our approach can be utilized with various distances which utilize covariance matrix structures. For instance, if we con￾sider Σ as the identity matrix I, which corr…
Figure 9
Figure 9. Figure 9: Visualization of the phenomenon that Mahalanobis dis￾tance may become less sensitive to OOD samples that align with the directions of high variance in ID data caused by outliers. The bold (solid) line is the geodesic line without manually added out￾lier points, and the…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 34 canonical work pages

  1. [1]

    H., Park, G.-M., and Kim, S

    Ahn, Y. H., Park, G.-M., and Kim, S. T. Line: Out-of-distribution detection by leveraging important neurons. In 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp.\ 19852--19862. IEEE, 2023

    work page 2023

  2. [2]

    B., Belkhir, N., Popescu, S., Manzanera, A., and Franchi, G

    Ammar, M. B., Belkhir, N., Popescu, S., Manzanera, A., and Franchi, G. NECO : NE ural collapse based out-of-distribution detection. In The Twelfth International Conference on Learning Representations, 2024. URL https://openreview.net/forum?id=9ROuKblmi7

    work page 2024

  3. [3]

    Scaling out-of-distribution detection for real-world settings

    Basart, S., Mantas, M., Mohammadreza, M., Jacob, S., and Dawn, S. Scaling out-of-distribution detection for real-world settings. In International Conference on Machine Learning, 2022

    work page 2022

  4. [4]

    L., Li, X., He, W., Gou, L., and Ren, L

    Behpour, S., Doan, T. L., Li, X., He, W., Gou, L., and Ren, L. Gradorth: A simple yet efficient out-of-distribution detection with orthogonal projection of gradients. Advances in Neural Information Processing Systems, 36: 0 38206--38230, 2023

    work page 2023

  5. [5]

    Emerging properties in self-supervised vision transformers

    Caron, M., Touvron, H., Misra, I., J\'egou, H., Mairal, J., Bojanowski, P., and Joulin, A. Emerging properties in self-supervised vision transformers. In Proceedings of the International Conference on Computer Vision (ICCV), 2021

    2021

  6. [6]

    Atom: Robustifying out-of-distribution detection using outlier mining

    Chen, J., Li, Y., Wu, X., Liang, Y., and Jha, S. Atom: Robustifying out-of-distribution detection using outlier mining. In Machine Learning and Knowledge Discovery in Databases. Research Track: European Conference, ECML PKDD 2021, Bilbao, Spain, September 13--17, 2021, Proceedings, Part III 21, pp.\ 430--445. Springer, 2021

    work page 2021

  7. [7]

    Chen, Y., Lin, Y., Xu, R., and Vela, P. A. Wdiscood: Out-of-distribution detection via whitened linear discriminant analysis. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pp.\ 5298--5307, 2023

    work page 2023

  8. [8]

    Describing textures in the wild

    Cimpoi, M., Maji, S., Kokkinos, I., Mohamed, S., and Vedaldi, A. Describing textures in the wild. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp.\ 3606--3613, 2014

    work page 2014

Show all 55 references
  1. [9]

    Imagenet: A large-scale hierarchical image database

    Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., and Fei-Fei, L. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pp.\ 248--255. Ieee, 2009

    2009

  2. [10]

    and Taylor, G

    DeVries, T. and Taylor, G. W. Learning confidence for out-of-distribution detection in neural networks. arXiv preprint arXiv:1802.04865, 2018

    2018 arXiv

  3. [11]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., and Houlsby, N. An image is worth 16x16 words: Transformers for image recognition at scale. In International Conference on...

    2021

  4. [12]

    Grood: Gradient-aware out-of-distribution detection in interpolated manifolds

    ElAraby, M., Sahoo, S., Pequignot, Y., Novello, P., and Paull, L. Grood: Gradient-aware out-of-distribution detection in interpolated manifolds. arXiv preprint arXiv:2312.14427, 2023

    2023

  5. [13]

    Can autonomous vehicles identify, recover from, and adapt to distribution shifts? In International Conference on Machine Learning, pp.\ 3145--3153

    Filos, A., Tigkas, P., McAllister, R., Rhinehart, N., Levine, S., and Gal, Y. Can autonomous vehicles identify, recover from, and adapt to distribution shifts? In International Conference on Machine Learning, pp.\ 3145--3153. PMLR, 2020

    2020

  6. [14]

    Evidence reconciled neural network for out-of-distribution detection in medical images

    Fu, W., Chen, Y., Liu, W., Yue, X., and Ma, C. Evidence reconciled neural network for out-of-distribution detection in medical images. In Medical Image Computing and Computer Assisted Intervention -- MICCAI 2023, pp.\ 305--315. Springer Nature Switzerland, 2023. ISBN 978-3-031-43898-1

    2023

  7. [15]

    Haas, J., Yolland, W., and Rabus, B. T. Linking neural collapse and l2 normalization with improved out-of-distribution detection in deep neural networks. Transactions on Machine Learning Research, 2023. ISSN 2835-8856. URL https://openreview.net/forum?id=fjkN5Ur2d6

    2023

  8. [16]

    A survey on vision transformer

    Han, K., Wang, Y., Chen, H., Chen, X., Guo, J., Liu, Z., Tang, Y., Xiao, A., Xu, C., Xu, Y., et al. A survey on vision transformer. IEEE transactions on pattern analysis and machine intelligence, 45 0 (1): 0 87--110, 2022

    2022

  9. [17]

    Deep residual learning for image recognition

    He, K., Zhang, X., Ren, S., and Sun, J. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp.\ 770--778, 2016

    2016

  10. [18]

    Why relu networks yield high-confidence predictions far away from the training data and how to mitigate the problem

    Hein, M., Andriushchenko, M., and Bitterwolf, J. Why relu networks yield high-confidence predictions far away from the training data and how to mitigate the problem. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp.\ 41--50, 2019

    2019

  11. [19]

    and Gimpel, K

    Hendrycks, D. and Gimpel, K. A baseline for detecting misclassified and out-of-distribution examples in neural networks. In International Conference on Learning Representations, 2017. URL https://openreview.net/forum?id=Hkg4TI9xl

    2017

  12. [20]

    Natural adversarial examples

    Hendrycks, D., Zhao, K., Basart, S., Steinhardt, J., and Song, D. Natural adversarial examples. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp.\ 15262--15271, 2021

    2021

  13. [21]

    Pixmix: Dreamlike pictures comprehensively improve safety measures

    Hendrycks, D., Zou, A., Mazeika, M., Tang, L., Li, B., Song, D., and Steinhardt, J. Pixmix: Dreamlike pictures comprehensively improve safety measures. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.\ 16783--16792, 2022

    2022

  14. [22]

    and Li, Y

    Huang, R. and Li, Y. Mos: Towards scaling out-of-distribution detection for large semantic space. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.\ 8710--8719, 2021

    2021

  15. [23]

    On the importance of gradients for detecting distributional shifts in the wild

    Huang, R., Geng, A., and Li, Y. On the importance of gradients for detecting distributional shifts in the wild. Advances in Neural Information Processing Systems, 34: 0 677--689, 2021

    2021

  16. [24]

    Enhancing camera surveillance using computer vision: a research note

    Idrees, H., Shah, M., and Surette, R. Enhancing camera surveillance using computer vision: a research note. Policing: An International Journal, 41 0 (2): 0 292--307, 2018

    2018

  17. [25]

    Openimages: A public dataset for large-scale multi-label and multi-class image classification

    Krasin, I., Duerig, T., Alldrin, N., Ferrari, V., Abu-El-Haija, S., Kuznetsova, A., Rom, H., Uijlings, J., Popov, S., Veit, A., Belongie, S., Gomes, V., Gupta, A., Sun, C., Chechik, G., Cai, D., Feng, Z., Narayanan, D., and Murphy, K. Openimages: A public dataset for large-sca...

    2017

  18. [26]

    Crafting papers on machine learning

    Langley, P. Crafting papers on machine learning. In Langley, P. (ed.), Proceedings of the 17th International Conference on Machine Learning (ICML 2000), pp.\ 1207--1216, Stanford, CA, 2000. Morgan Kaufmann

    2000

  19. [27]

    A simple unified framework for detecting out-of-distribution samples and adversarial attacks

    Lee, K., Lee, K., Lee, H., and Shin, J. A simple unified framework for detecting out-of-distribution samples and adversarial attacks. Advances in neural information processing systems, 31, 2018

    2018

  20. [28]

    Enhancing the reliability of out-of-distribution image detection in neural networks

    Liang, S., Li, Y., and Srikant, R. Enhancing the reliability of out-of-distribution image detection in neural networks. In International Conference on Learning Representations, 2018. URL https://openreview.net/forum?id=H1VGkIxRZ

    2018

  21. [29]

    and Qin, Y

    Liu, L. and Qin, Y. Fast decision boundary based out-of-distribution detector. arXiv preprint arXiv:2312.11536, 2023

    2023 arXiv

  22. [30]

    Energy-based out-of-distribution detection

    Liu, W., Wang, X., Owens, J., and Li, Y. Energy-based out-of-distribution detection. Advances in neural information processing systems, 33: 0 21464--21475, 2020

    2020

  23. [31]

    and Li, Y

    Morteza, P. and Li, Y. Provable guarantees for understanding out-of-distribution detection. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pp.\ 7831--7840, 2022

    2022

  24. [32]

    Y., et al

    Netzer, Y., Wang, T., Coates, A., Bissacco, A., Wu, B., Ng, A. Y., et al. Reading digits in natural images with unsupervised feature learning. In NIPS workshop on deep learning and unsupervised feature learning, volume 2011, pp.\ 4. Granada, 2011

    2011

  25. [33]

    Boundary-aware backward-compatible representation via adversarial learning in image retrieval

    Pan, T., Xu, F., Yang, X., He, S., Jiang, C., Guo, Q., Qian, F., Zhang, X., Cheng, Y., Yang, L., et al. Boundary-aware backward-compatible representation via adversarial learning in image retrieval. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog...

    2023

  26. [34]

    Papyan, V., Han, X., and Donoho, D. L. Prevalence of neural collapse during the terminal phase of deep learning training. Proceedings of the National Academy of Sciences, 117 0 (40): 0 24652--24663, 2020

    2020

  27. [35]

    Conjnorm: Tractable density estimation for out-of-distribution detection

    Peng, B., Luo, Y., Zhang, Y., Li, Y., and Fang, Z. Conjnorm: Tractable density estimation for out-of-distribution detection. In The Twelfth International Conference on Learning Representations, 2024. URL https://openreview.net/forum?id=1pSL2cXWoz

    2024

  28. [36]

    I., and Strapasson, J

    Pinele, J., Costa, S. I., and Strapasson, J. E. On the fisher-rao information metric in the space of normal distributions. In International Conference on Geometric Science of Information, pp.\ 676--684. Springer, 2019

    2019

  29. [37]

    G., Padhy, S., and Lakshminarayanan, B

    Ren, J., Fort, S., Liu, J., Roy, A. G., Padhy, S., and Lakshminarayanan, B. A simple fix to mahalanobis distance for improving near-ood detection. arXiv preprint arXiv:2106.09022, 2021

    2021 arXiv

  30. [38]

    \ SSD \ : A unified framework for self-supervised outlier detection

    Sehwag, V., Chiang, M., and Mittal, P. \ SSD \ : A unified framework for self-supervised outlier detection. In International Conference on Learning Representations, 2021. URL https://openreview.net/forum?id=v5gjXpmR8J

    2021

  31. [39]

    and Li, Y

    Sun, Y. and Li, Y. Dice: Leveraging sparsification for out-of-distribution detection. In European Conference on Computer Vision, pp.\ 691--708. Springer, 2022

    2022

  32. [40]

    React: Out-of-distribution detection with rectified activations

    Sun, Y., Guo, C., and Li, Y. React: Out-of-distribution detection with rectified activations. Advances in Neural Information Processing Systems, 34: 0 144--157, 2021

    2021

  33. [41]

    Out-of-distribution detection with deep nearest neighbors

    Sun, Y., Ming, Y., Zhu, X., and Li, Y. Out-of-distribution detection with deep nearest neighbors. In International Conference on Machine Learning, pp.\ 20827--20840. PMLR, 2022

    2022

  34. [42]

    M., and Liang, P

    Tajwar, F., Kumar, A., Xie, S. M., and Liang, P. No true state-of-the-art? ood detection methods are inconsistent across datasets. arXiv preprint arXiv:2109.05554, 2021

    2021 arXiv

  35. [43]

    Training data-efficient image transformers & distillation through attention

    Touvron, H., Cord, M., Douze, M., Massa, F., Sablayrolles, A., and J \'e gou, H. Training data-efficient image transformers & distillation through attention. In International conference on machine learning, pp.\ 10347--10357. PMLR, 2021

    2021

  36. [44]

    The inaturalist species classification and detection dataset

    Van Horn, G., Mac Aodha, O., Song, Y., Cui, Y., Sun, C., Shepard, A., Adam, H., Perona, P., and Belongie, S. The inaturalist species classification and detection dataset. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp.\ 8769--8778, 2018

    2018

  37. [45]

    Vyas, A., Jammalamadaka, N., Zhu, X., Das, D., Kaul, B., and Willke, T. L. Out-of-distribution detection using an ensemble of self supervised leave-out classifiers. In Proceedings of the European conference on computer vision (ECCV), pp.\ 550--564, 2018

    2018

  38. [46]

    Vim: Out-of-distribution with virtual-logit matching

    Wang, H., Li, Z., Feng, L., and Zhang, W. Vim: Out-of-distribution with virtual-logit matching. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp.\ 4921--4930, 2022

    2022

  39. [47]

    Energy-based open-world uncertainty modeling for confidence calibration

    Wang, Y., Li, B., Che, T., Zhou, K., Liu, Z., and Li, D. Energy-based open-world uncertainty modeling for confidence calibration. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pp.\ 9302--9311, 2021

    2021

  40. [48]

    A., Oliva, A., and Torralba, A

    Xiao, J., Hays, J., Ehinger, K. A., Oliva, A., and Torralba, A. Sun database: Large-scale scene recognition from abbey to zoo. In 2010 IEEE computer society conference on computer vision and pattern recognition, pp.\ 3485--3492. IEEE, 2010

    2010

  41. [49]

    Scaling for training time and post-hoc out-of-distribution detection enhancement

    Xu, K., Chen, R., Franchi, G., and Yao, A. Scaling for training time and post-hoc out-of-distribution detection enhancement. In The Twelfth International Conference on Learning Representations, 2024. URL https://openreview.net/forum?id=RDSTjtnqCg

    2024

  42. [50]

    A., Zhang, Y., Finkelstein, A., Kulkarni, S

    Xu, P., Ehinger, K. A., Zhang, Y., Finkelstein, A., Kulkarni, S. R., and Xiao, J. Turkergaze: Crowdsourcing saliency with webcam based eye tracking. arXiv preprint arXiv:1504.06755, 2015

    2015 arXiv

  43. [51]

    Openood: Benchmarking generalized out-of-distribution detection

    Yang, J., Wang, P., Zou, D., Zhou, Z., Ding, K., Peng, W., Wang, H., Chen, G., Li, B., Sun, Y., et al. Openood: Benchmarking generalized out-of-distribution detection. Advances in Neural Information Processing Systems, 35: 0 32598--32611, 2022

    2022

  44. [52]

    Generalized out-of-distribution detection: A survey

    Yang, J., Zhou, K., Li, Y., and Liu, Z. Generalized out-of-distribution detection: A survey. International Journal of Computer Vision, pp.\ 1--28, 2024

    2024

  45. [53]

    Lsun: Construction of a large-scale image dataset using deep learning with humans in the loop

    Yu, F., Seff, A., Zhang, Y., Song, S., Funkhouser, T., and Xiao, J. Lsun: Construction of a large-scale image dataset using deep learning with humans in the loop. arXiv preprint arXiv:1506.03365, 2015

    2015 arXiv

  46. [54]

    Places: A 10 million image database for scene recognition

    Zhou, B., Lapedriza, A., Khosla, A., Oliva, A., and Torralba, A. Places: A 10 million image database for scene recognition. IEEE transactions on pattern analysis and machine intelligence, 40 0 (6): 0 1452--1464, 2017

    2017

  47. [55]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.