pith. sign in

arxiv: 2605.20728 · v1 · pith:VHGYIL7Rnew · submitted 2026-05-20 · 💻 cs.CV

Early High-Frequency Injection for Geometry-Sensitive OOD Detection

Chuanjie Cheng , Ningkang Peng , Chenxi Liu , Yifan He , Peirong Ma , Yanhui Gu This is my paper

Pith reviewed 2026-05-21 05:09 UTC · model grok-4.3

classification 💻 cs.CV
keywords out-of-distribution detectionhigh-frequency injectionfeature geometryMahalanobis scoringOOD detectionrepresentation learningCIFAR-100ImageNet-100
0
0 comments X

The pith

Injecting high-frequency input components early reshapes neural features to separate in-distribution from out-of-distribution samples more cleanly.

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

The paper starts from the observation that post-hoc OOD detectors succeed only when the learned representation already encodes useful geometry. A band-wise MMD squared diagnostic across several training methods shows that low-frequency input bands produce weak ID/OOD discrepancy while higher-frequency bands produce stronger separability. This finding motivates an input-side change, EIHF, that adds high-frequency evidence before the first convolution while leaving the training objective unchanged. Under matched conditions the intervention reduces overlap in Mahalanobis scores and yields lower FPR95 on CIFAR-100 and the best average FPR95 on ImageNet-100. The approach shows a clear limitation when the shift is scene-centric rather than object-centric.

Core claim

Under matched training and scoring settings, early high-frequency injection (EIHF) reshapes class-conditional feature geometry and reduces ID/OOD Mahalanobis score overlap. The method works by exposing higher-frequency bands at the input before the first convolution, exploiting the observation that these bands induce stronger feature discrepancy than low-frequency bands.

What carries the argument

EIHF, an input-side intervention that adds high-frequency evidence before the first convolution to reshape class-conditional feature geometry for geometry-sensitive OOD scoring.

If this is right

  • Geometry-sensitive post-hoc detectors such as Mahalanobis distance scoring receive the largest gains.
  • Performance improves on object-centric shifts such as CIFAR-100 and ImageNet-100.
  • No change to the training loss or architecture is required.
  • A performance drop appears on scene-centric shifts such as Places.

Where Pith is reading between the lines

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

  • The same frequency-band analysis could be applied to other representation-learning objectives to decide where to inject information.
  • Input-level frequency interventions may complement or replace some post-training scoring refinements.
  • The limitation on scene-centric data suggests that frequency content interacts with the semantic granularity of the shift.

Load-bearing premise

The band-wise MMD squared diagnostic correctly identifies higher-frequency bands as carrying stronger ID/OOD separability that early injection can exploit.

What would settle it

Measuring the Mahalanobis score distributions on a held-out ID/OOD pair after applying EIHF; if the overlap between the two distributions does not decrease relative to the baseline, the central claim is false.

Figures

Figures reproduced from arXiv: 2605.20728 by Chenxi Liu, Chuanjie Cheng, Ningkang Peng, Peirong Ma, Yanhui Gu, Yifan He.

Figure 1
Figure 1. Figure 1: Band-wise MMD2 with CIFAR-100 as ID. The diagnostic is computed on fixed encoders by feeding band-limited inputs and measuring final-feature discrepancy. Low-frequency bands show weaker ID/OOD feature discrepancy, while mid- and high-frequency bands show stronger separability across OOD datasets and training objectives. for OOD detection unequally accessible. Since low-frequency structure can be shared by … view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the EIHF framework. EIHF computes a fixed high-frequency residual from [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Mechanism analysis of how EIHF shapes representation geometry. (a) ID distances to class centers; EIHF reduces the mean distance by 20.9%. (b) ID and OOD Mahalanobis score distributions; EIHF reduces score overlap near the FPR95 operating point. Experiments use ResNet￾34 on CIFAR-100. 3.3 EIHF: Early High-Frequency Injection Guided by the band-wise diagnostic, we instantiate the input transformation η as E… view at source ↗
Figure 4
Figure 4. Figure 4: ID classification accuracy on CIFAR-100, comparing EIHF with the representation-learning [PITH_FULL_IMAGE:figures/full_fig_p013_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Mahalanobis score overlap versus average FPR95 on CIFAR-100. Lower overlap corre [PITH_FULL_IMAGE:figures/full_fig_p014_5.png] view at source ↗
read the original abstract

Post-hoc OOD detectors score logits or features after training, so their success depends on the geometry already encoded in the representation. We revisit this assumption through a band-wise MMD^2 analysis across CE, SimCLR, SupCon, and the OOD-oriented representation method PALM. In our diagnostic, low-frequency input bands induce weaker ID/OOD feature discrepancy, whereas higher-frequency bands tend to provide stronger separability. This observation motivates EIHF, an input-side intervention that exposes high-frequency evidence before the first convolution without changing the training objective. EIHF is strongest for geometry-sensitive OOD detection: under matched training and scoring settings, it reshapes class-conditional feature geometry and reduces ID/OOD Mahalanobis score overlap. Experiments on CIFAR-100 and ImageNet-100 show gains on CIFAR-100 and the best average FPR95 with second-best average AUROC on ImageNet-100, while also revealing a limitation on the scene-centric Places shift. Code is available at https://anonymous.4open.science/r/EIHF.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The paper claims that a band-wise MMD² analysis across CE, SimCLR, SupCon, and PALM representations shows weaker ID/OOD separability in low-frequency input bands and stronger separability in higher-frequency bands. This motivates EIHF, an input-side intervention that injects high-frequency evidence before the first convolution without altering the training objective. Under matched training and scoring, EIHF reshapes class-conditional feature geometry, reduces Mahalanobis ID/OOD overlap, and yields gains on CIFAR-100 plus the best average FPR95 on ImageNet-100, while noting a limitation on the Places shift. Code is released.

Significance. If the central claim holds after addressing the diagnostic, EIHF provides a lightweight, training-agnostic way to improve geometry-sensitive post-hoc OOD detectors by directly influencing early feature geometry. The empirical results on standard benchmarks, code availability, and frequency-based diagnostic add practical and conceptual value to the OOD literature.

major comments (1)
  1. [Band-wise MMD² analysis (motivation)] The band-wise MMD² diagnostic may confound frequency content with per-band energy. Higher-frequency bands inherently carry lower amplitude; without explicit per-band energy or variance normalization before feature extraction, any MMD² increase could trace to signal-strength differences rather than frequency-specific discriminative structure. This assumption is load-bearing for motivating the early-injection intervention (see Abstract and motivation section).
minor comments (2)
  1. [Abstract] The abstract and methods lack precise implementation details on the high-frequency injection (filtering method, scaling, exact masking).
  2. [Experiments] Statistical significance of reported gains should be included, along with expanded controls or analysis for the Places limitation.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback and for highlighting a potential confound in our band-wise MMD² diagnostic. We address this point directly below and have revised the manuscript to strengthen the motivation for EIHF.

read point-by-point responses

  1. Referee: [Band-wise MMD² analysis (motivation)] The band-wise MMD² diagnostic may confound frequency content with per-band energy. Higher-frequency bands inherently carry lower amplitude; without explicit per-band energy or variance normalization before feature extraction, any MMD² increase could trace to signal-strength differences rather than frequency-specific discriminative structure. This assumption is load-bearing for motivating the early-injection intervention (see Abstract and motivation section).

    Authors: We agree that the absence of per-band energy normalization represents a valid concern, as higher-frequency bands do carry lower amplitude on average and this could partially drive observed MMD² trends. In the original analysis we did not apply explicit per-band variance normalization before feature extraction. To resolve this, we have revised the diagnostic: each frequency band is now normalized by its own standard deviation prior to MMD² computation. The updated results (new Figure 2 and expanded Section 3.2) preserve the key pattern—low-frequency bands continue to yield weaker ID/OOD separability while higher-frequency bands yield stronger separability—across CE, SimCLR, SupCon, and PALM representations. We have also added a brief discussion of this normalization step in the motivation section to make the frequency-specific claim more robust. These changes directly address the load-bearing assumption for EIHF. revision: yes

Circularity Check

0 steps flagged

No circularity: independent diagnostic motivates intervention with empirical validation

full rationale

The derivation begins with a band-wise MMD^2 analysis across multiple training methods (CE, SimCLR, SupCon, PALM) that identifies frequency-dependent ID/OOD separability in features. This observation directly motivates the EIHF input intervention, which is then evaluated by measuring changes in class-conditional geometry and Mahalanobis score overlap on CIFAR-100 and ImageNet-100. No equations reduce the claimed gains to a fitted parameter or self-referential definition; the central result is the measured effect of the intervention under matched settings. No load-bearing self-citations or uniqueness theorems are invoked. The chain is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach rests on an empirical frequency-separability observation rather than new axioms or invented entities.

axioms (1)
  • domain assumption Higher-frequency input bands induce stronger ID/OOD feature discrepancy than low-frequency bands
    Derived from the band-wise MMD^2 analysis across CE, SimCLR, SupCon, and PALM

pith-pipeline@v0.9.0 · 5732 in / 1130 out tokens · 32681 ms · 2026-05-21T05:09:45.589112+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

57 extracted references · 57 canonical work pages · 4 internal anchors

  1. [1]

    Invariant scattering convolution networks

    Joan Bruna and Stephane Mallat. Invariant scattering convolution networks. IEEE Transactions on Pattern Analysis and Machine Intelligence , 35(8):1872--1886, 2013

    work page 2013

  2. [2]

    A computational approach to edge detection

    John Canny. A computational approach to edge detection. IEEE Transactions on Pattern Analysis and Machine Intelligence , 8(6):679--698, 1986

    work page 1986

  3. [3]

    Out-of-distribution detection via frequency-regularized generative models

    Mu Cai and Yixuan Li. Out-of-distribution detection via frequency-regularized generative models. In IEEE/CVF Winter Conference on Applications of Computer Vision , 2023

    work page 2023

  4. [4]

    Deep clustering for unsupervised learning of visual features

    Mathilde Caron, Piotr Bojanowski, Armand Joulin, and Matthijs Douze. Deep clustering for unsupervised learning of visual features. In European Conference on Computer Vision , 2018

    work page 2018

  5. [5]

    Unsupervised learning of visual features by contrasting cluster assignments

    Mathilde Caron, Ishan Misra, Julien Mairal, Priya Goyal, Piotr Bojanowski, and Armand Joulin. Unsupervised learning of visual features by contrasting cluster assignments. In Advances in Neural Information Processing Systems , 2020

    work page 2020

  6. [6]

    A simple framework for contrastive learning of visual representations

    Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. A simple framework for contrastive learning of visual representations. In International Conference on Machine Learning , 2020

    work page 2020

  7. [7]

    Describing textures in the wild

    Mircea Cimpoi, Subhransu Maji, Iasonas Kokkinos, Sammy Mohamed, and Andrea Vedaldi. Describing textures in the wild. In IEEE Conference on Computer Vision and Pattern Recognition , 2014

    work page 2014

  8. [8]

    ImageNet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. ImageNet: A large-scale hierarchical image database. In IEEE Conference on Computer Vision and Pattern Recognition , 2009

    work page 2009

  9. [9]

    Extremely simple activation shaping for out-of-distribution detection

    Andrija Djurisic, Nebojsa Bozanic, Arjun Ashok, and Rosanne Liu. Extremely simple activation shaping for out-of-distribution detection. In International Conference on Learning Representations , 2023

    work page 2023

  10. [10]

    VOS: Learning what you don't know by virtual outlier synthesis

    Xuefeng Du, Zhaoning Wang, Mu Cai, and Yixuan Li. VOS: Learning what you don't know by virtual outlier synthesis. In International Conference on Learning Representations , 2022

    work page 2022

  11. [11]

    The open world assumption

    Nick Drummond and Rob Shearer. The open world assumption. In The Closed World of Databases Meets the Open World of the Semantic Web , 2006

    work page 2006

  12. [12]

    Wichmann, and Wieland Brendel

    Robert Geirhos, Patricia Rubisch, Claudio Michaelis, Matthias Bethge, Felix A. Wichmann, and Wieland Brendel. ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness. In International Conference on Learning Representations , 2019

    work page 2019

  13. [13]

    Your classifier is secretly an energy based model and you should treat it like one

    Will Grathwohl, Kuan-Chieh Wang, Joern-Henrik Jacobsen, David Duvenaud, Mohammad Norouzi, and Kevin Swersky. Your classifier is secretly an energy based model and you should treat it like one. In International Conference on Learning Representations , 2020

    work page 2020

  14. [14]

    Bootstrap your own latent: A new approach to self-supervised learning

    Jean-Bastien Grill, Florian Strub, Florent Altche, Corentin Tallec, Pierre Richemond, Elena Buchatskaya, Carl Doersch, Bernardo Avila Pires, Zhaohan Guo, and others. Bootstrap your own latent: A new approach to self-supervised learning. In Advances in Neural Information Processing Systems , 2020

    work page 2020

  15. [15]

    Borgwardt, Malte J

    Arthur Gretton, Karsten M. Borgwardt, Malte J. Rasch, Bernhard Schoelkopf, and Alexander Smola. A kernel two-sample test. Journal of Machine Learning Research , 13:723--773, 2012

    work page 2012

  16. [16]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In IEEE Conference on Computer Vision and Pattern Recognition , 2016

    work page 2016

  17. [17]

    Momentum contrast for unsupervised visual representation learning

    Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. Momentum contrast for unsupervised visual representation learning. In IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020

    work page 2020

  18. [18]

    A baseline for detecting misclassified and out-of-distribution examples in neural networks

    Dan Hendrycks and Kevin Gimpel. A baseline for detecting misclassified and out-of-distribution examples in neural networks. In International Conference on Learning Representations , 2017

    work page 2017

  19. [19]

    A survey of safety and trustworthiness of deep neural networks

    Xiaowei Huang, Daniel Kroening, Wenjie Ruan, James Sharp, Youcheng Sun, Emese Thamo, Min Wu, and Xinping Yi. A survey of safety and trustworthiness of deep neural networks. Computer Science Review , 37:100270, 2020

    work page 2020

  20. [20]

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

    Rui Huang, Andrew Geng, and Yixuan Li. On the importance of gradients for detecting distributional shifts in the wild. In Advances in Neural Information Processing Systems , 2021

    work page 2021

  21. [21]

    Measuring the tendency of CNNs to Learn Surface Statistical Regularities

    Jason Jo and Yoshua Bengio. Measuring the tendency of CNNs to learn surface statistical regularities. arXiv preprint arXiv:1711.11561, 2017

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  22. [22]

    Supervised contrastive learning

    Prannay Khosla, Piotr Teterwak, Chen Wang, Aaron Sarna, Yonglong Tian, Phillip Isola, Aaron Maschinot, Ce Liu, and Dilip Krishnan. Supervised contrastive learning. In Advances in Neural Information Processing Systems , 2020

    work page 2020

  23. [23]

    Kingma and Prafulla Dhariwal

    Durk P. Kingma and Prafulla Dhariwal. Glow: Generative flow with invertible 1x1 convolutions. In Advances in Neural Information Processing Systems , 2018

    work page 2018

  24. [24]

    Learning multiple layers of features from tiny images

    Alex Krizhevsky. Learning multiple layers of features from tiny images. Technical report, University of Toronto, 2009

    work page 2009

  25. [25]

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

    Kimin Lee, Kibok Lee, Honglak Lee, and Jinwoo Shin. A simple unified framework for detecting out-of-distribution samples and adversarial attacks. In Advances in Neural Information Processing Systems , 2018

    work page 2018

  26. [26]

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

    Shiyu Liang, Yixuan Li, and Rayadurgam Srikant. Enhancing the reliability of out-of-distribution image detection in neural networks. In International Conference on Learning Representations , 2018

    work page 2018

  27. [27]

    Junnan Li, Pan Zhou, Caiming Xiong, and Steven C. H. Hoi. Prototypical contrastive learning of unsupervised representations. In International Conference on Learning Representations , 2021

    work page 2021

  28. [28]

    Owens, and Yixuan Li

    Weitang Liu, Xiaoyun Wang, John D. Owens, and Yixuan Li. Energy-based out-of-distribution detection. In Advances in Neural Information Processing Systems , 2020

    work page 2020

  29. [29]

    Learning with mixture of prototypes for out-of-distribution detection

    Haodong Lu, Dong Gong, Shuo Wang, Jason Xue, Lina Yao, and Kristen Moore. Learning with mixture of prototypes for out-of-distribution detection. arXiv preprint arXiv:2402.02653, 2024

    work page arXiv 2024

  30. [30]

    A theory for multiresolution signal decomposition: The wavelet representation

    Stephane Mallat. A theory for multiresolution signal decomposition: The wavelet representation. IEEE Transactions on Pattern Analysis and Machine Intelligence , 11(7):674--693, 1989

    work page 1989

  31. [31]

    How to exploit hyperspherical embeddings for out-of-distribution detection? In International Conference on Learning Representations , 2023

    Yifei Ming, Yiyou Sun, Ousmane Dia, and Yixuan Li. How to exploit hyperspherical embeddings for out-of-distribution detection? In International Conference on Learning Representations , 2023

    work page 2023

  32. [32]

    Yuval Netzer, Tao Wang, Adam Coates, Alessandro Bissacco, Bo Wu, and Andrew Y. Ng. Reading digits in natural images with unsupervised feature learning. In NeurIPS Workshop on Deep Learning and Unsupervised Feature Learning , 2011

    work page 2011

  33. [33]

    Deep neural networks are easily fooled: High confidence predictions for unrecognizable images

    Anh Nguyen, Jason Yosinski, and Jeff Clune. Deep neural networks are easily fooled: High confidence predictions for unrecognizable images. In IEEE Conference on Computer Vision and Pattern Recognition , 2015

    work page 2015

  34. [34]

    Representation Learning with Contrastive Predictive Coding

    Aaron van den Oord, Yazhe Li, and Oriol Vinyals. Representation learning with contrastive predictive coding. arXiv preprint arXiv:1807.03748, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  35. [35]

    Distributional prototype learning for out-of-distribution detection

    Bo Peng, Jie Lu, Yonggang Zhang, Guangquan Zhang, and Zhen Fang. Distributional prototype learning for out-of-distribution detection. In ACM SIGKDD Conference on Knowledge Discovery and Data Mining , 2025

    work page 2025

  36. [36]

    Hamprecht, Yoshua Bengio, and Aaron Courville

    Nasim Rahaman, Aristide Baratin, Devansh Arpit, Felix Draxler, Min Lin, Fred A. Hamprecht, Yoshua Bengio, and Aaron Courville. On the spectral bias of neural networks. In International Conference on Machine Learning , 2019

    work page 2019

  37. [37]

    Liu, Emily Fertig, Jasper Snoek, Ryan Poplin, Mark DePristo, Joshua Dillon, and Balaji Lakshminarayanan

    Jie Ren, Peter J. Liu, Emily Fertig, Jasper Snoek, Ryan Poplin, Mark DePristo, Joshua Dillon, and Balaji Lakshminarayanan. Likelihood ratios for out-of-distribution detection. In Advances in Neural Information Processing Systems , 2019

    work page 2019

  38. [38]

    Contrastive learning with hard negative samples

    Joshua David Robinson, Ching-Yao Chuang, Suvrit Sra, and Stefanie Jegelka. Contrastive learning with hard negative samples. In International Conference on Learning Representations , 2021

    work page 2021

  39. [39]

    SSD: A unified framework for self-supervised outlier detection

    Vikash Sehwag, Mung Chiang, and Prateek Mittal. SSD: A unified framework for self-supervised outlier detection. In International Conference on Learning Representations , 2021

    work page 2021

  40. [40]

    Prototypical networks for few-shot learning

    Jake Snell, Kevin Swersky, and Richard Zemel. Prototypical networks for few-shot learning. In Advances in Neural Information Processing Systems , 2017

    work page 2017

  41. [41]

    ReAct: Out-of-distribution detection with rectified activations

    Yiyou Sun, Chuan Guo, and Yixuan Li. ReAct: Out-of-distribution detection with rectified activations. In Advances in Neural Information Processing Systems , 2021

    work page 2021

  42. [42]

    Out-of-distribution detection with deep nearest neighbors

    Yiyou Sun, Yifei Ming, Xiaojin Zhu, and Yixuan Li. Out-of-distribution detection with deep nearest neighbors. In International Conference on Machine Learning , 2022

    work page 2022

  43. [43]

    CSI: Novelty detection via contrastive learning on distributionally shifted instances

    Jihoon Tack, Sangwoo Mo, Jongheon Jeong, and Jinwoo Shin. CSI: Novelty detection via contrastive learning on distributionally shifted instances. In Advances in Neural Information Processing Systems , 2020

    work page 2020

  44. [44]

    Non-parametric outlier synthesis

    Leitian Tao, Xuefeng Du, Jerry Zhu, and Yixuan Li. Non-parametric outlier synthesis. In International Conference on Learning Representations , 2023

    work page 2023

  45. [45]

    Contrastive multiview coding

    Yonglong Tian, Dilip Krishnan, and Phillip Isola. Contrastive multiview coding. In European Conference on Computer Vision , 2020

    work page 2020

  46. [46]

    Energy-based open-world uncertainty modeling for confidence calibration

    Yezhen Wang, Bo Li, Tong Che, Kaiyang Zhou, Ziwei Liu, and Dongsheng Li. Energy-based open-world uncertainty modeling for confidence calibration. In IEEE/CVF International Conference on Computer Vision , 2021

    work page 2021

  47. [47]

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

    Haoqi Wang, Zhizhong Li, Litong Feng, and Wayne Zhang. ViM: Out-of-distribution with virtual-logit matching. In IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022

    work page 2022

  48. [48]

    Mitigating neural network overconfidence with logit normalization

    Hongxin Wei, Renchunzi Xie, Hao Cheng, Lei Feng, Bo An, and Yixuan Li. Mitigating neural network overconfidence with logit normalization. In International Conference on Machine Learning , 2022

    work page 2022

  49. [49]

    Yu, and Dahua Lin

    Zhirong Wu, Yuanjun Xiong, Stella X. Yu, and Dahua Lin. Unsupervised feature learning via non-parametric instance discrimination. In IEEE Conference on Computer Vision and Pattern Recognition , 2018

    work page 2018

  50. [50]

    Likelihood regret: An out-of-distribution detection score for variational auto-encoder

    Zhisheng Xiao, Qing Yan, and Yali Amit. Likelihood regret: An out-of-distribution detection score for variational auto-encoder. In Advances in Neural Information Processing Systems , 2020

    work page 2020

  51. [51]

    TurkerGaze: Crowdsourcing Saliency with Webcam based Eye Tracking

    Pingmei Xu, Krista A. Ehinger, Yinda Zhang, Adam Finkelstein, Sanjeev R. Kulkarni, and Jianxiong Xiao. TurkerGaze: Crowdsourcing saliency with webcam based eye tracking. arXiv preprint arXiv:1504.06755, 2015

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  52. [52]

    Frequency principle: Fourier analysis sheds light on deep neural networks

    Zhi-Qin John Xu, Yaoyu Zhang, Tao Luo, Yanyang Xiao, and Zheng Ma. Frequency principle: Fourier analysis sheds light on deep neural networks. Communications in Computational Physics , 28(5):1746--1767, 2020

    work page 2020

  53. [53]

    Generalized out-of-distribution detection: A survey

    Jingkang Yang, Kaiyang Zhou, Yixuan Li, and Ziwei Liu. Generalized out-of-distribution detection: A survey. International Journal of Computer Vision , 132:5635--5662, 2024

    work page 2024

  54. [54]

    Cubuk, and Justin Gilmer

    Dong Yin, Raphael Gontijo Lopes, Jonathon Shlens, Ekin D. Cubuk, and Justin Gilmer. A Fourier perspective on model robustness in computer vision. In Advances in Neural Information Processing Systems , 2019

    work page 2019

  55. [55]

    LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop

    Fisher Yu, Yinda Zhang, Shuran Song, Ari Seff, and Jianxiong Xiao. LSUN: Construction of a large-scale image dataset using deep learning with humans in the loop. arXiv preprint arXiv:1506.03365, 2015

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  56. [56]

    Learning to shape in-distribution feature space for out-of-distribution detection

    Yonggang Zhang, Jie Lu, Bo Peng, Zhen Fang, and Yiu-ming Cheung. Learning to shape in-distribution feature space for out-of-distribution detection. In Advances in Neural Information Processing Systems , 2024

    work page 2024

  57. [57]

    Places: A 10 million image database for scene recognition

    Bolei Zhou, Agata Lapedriza, Aditya Khosla, Aude Oliva, and Antonio Torralba. Places: A 10 million image database for scene recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence , 40(6):1452--1464, 2017

    work page 2017