betaFeedback

WorksComputer Science

Residual Calibration via Local Feature-Space Refinement

Abstract

Deep neural networks often struggle to provide reliable confidence estimates, making post-hoc calibration essential for reliable decision-making. Existing post-hoc methods largely rely on score-based transformations, which improve marginal calibration but often fail to capture spatially varying miscalibration in local regions of the learned feature space. To address this gap, we propose Local Adaptive Probability Estimation (LAPE), a post-hoc residual calibration framework. Specifically, LAPE treats the global calibrator output as a stable anchor and refines the individual prediction by leveraging non-parametric local evidence from neighboring calibration samples. As a plug-and-play refinement module, LAPE effectively utilizes the information about feature-space geometry and can reuse the calibration data that are used to construct the global calibrator. Furthermore, we extend LAPE to the setting with multi-view inputs through Cascade Feature Augmentation Fusion (CFAF). Extensive experiments demonstrate that the proposed methods can improve the calibration quality of various existing baselines and remain efficient and robust across different settings.

Introduction

Deep neural networks have exhibited remarkable accuracy on classification problems, but often fail to provide reliable confidence scores that reflect the ground truth correctness likelihoods [14]. This mismatch can appear as overconfidence or underconfidence, both of which are critical especially in high-stakes applications and downstream tasks relying prediction uncertainty [48 b]. For example, in automated clinical diagnostics, a reliable confidence score can indicate when a prediction should be deferred to experts for further examination, and in robust conformalized selection [48], a reliable confidence score can improve selection power. Therefore, a network should provide reliable calibrated probabilities rather than directly applying the point predictions [2, 25, 17].

Post-hoc calibration provides a cost-effective solution to rectify miscalibration without retraining. However, existing global parametric maps such as Temperature Scaling (TS) [14], ensemble TS (ETS) [52], and Dirichlet calibration [23] rely solely on scalar logits. Consequently, predictions with similar scores tend to obtain identical calibrations, ignoring the varying empirical reliability across distinct feature neighborhoods. To this end, local TS (LTS) [10] introduces spatial adaptivity, but its increased parametric capacity risks overfitting on limited calibration data [14, 23, 38]. Conversely, purely non-parametric local methods capture spatial patterns by adapting to neighborhood distributions but suffer from data inefficiency and discard informative global logits [50, 51, 29, 24]. These limitations raise a foundational challenge: how to develop a calibration framework that leverages local feature-space evidence to correct residual errors while strictly preserving global predictive stability?

To address the above challenge, we propose Local Adaptive Probability Estimation (LAPE), a post-hoc residual calibration framework that effectively utilizes local feature-level evidence to further enhance existing calibration methods. Specifically, LAPE formulates calibration as a residual refinement problem. It takes the output of any existing post-hoc calibrator as a stable anchor (also called global anchor), and then applies a local correction to it only when the nearby calibration samples support it. To tackle the overfitting issue caused by reusing the calibration set to fit both a global anchor and residual refinement, we introduce a VV-fold cross-fitting strategy. It generates out-of-sample anchor scores in the calibration set and can be more effective than the standard data-splitting strategy. Finally, we develop Cascade Feature Augmentation Fusion (CFAF), which extends LAPE to tackle multi-view scenarios [45, 4].

In theory, we establish the asymptotic consistency of the proposed LAPE. Empirically, we extensively evaluate LAPE across diverse benchmarks that encompass tabular and image-type data. Results show that LAPE helps to enhance the calibration performance of various baselines, such as TS [14], Feature Clipping (FC) [37], and Semantic-Aware Grouping (SAG) [46]. Extensive experiments in Appendix § 12 further demonstrate the efficiency, robustness, and effectiveness of LAPE.

We summarize our contributions as follows:

  1. We propose LAPE, a residual calibration framework that enhances existing post-hoc calibration methods by integrating feature-level evidence. Importantly, LAPE can reuse the calibration dataset for building the global calibrator and residual refinement.

  2. We extend LAPE to multi-view settings through CFAF, which combines calibrated view-level predictions with local refinement on a rectified joint feature space.

  3. We establish asymptotic consistency of LAPE, providing a consistency analysis for feature-space residual calibration of the full multiclass probability vector.

  4. We conduct comprehensive evaluations across various datasets and architectures, demonstrating that LAPE can improve the performance of existing post-hoc calibration methods.

Preliminary

Problem Setup

Let X∈XX \in\mathcal{X} denote the feature variable and Y∈Y={1,…,K}Y \in\mathcal{Y} = \{1,\ldots,K\} denote the label variable. Assume access to a calibration dataset {(xi,yi)}i=1n\{(x_i,y_i)\}_{i=1}^{n}. Let π^0:X→ΔK−1\hat{\pi}_0 : \mathcal{X} \to\Delta^{K-1} be a pre-trained soft-classifier for the KK-class classification, where ΔK−1\Delta^{K-1} represents the (K−1)(K-1)-dimensional probability simplex. Let p(k;x)=P(Y=k∣X=x)p(k;x) = \mathbb{P}(Y = k \mid X = x). For any x∈Xx \in\mathcal{X}, we write π^0(x)=(π^0(1;x),…,π^0(K;x))\hat{\pi}_0(x) = (\hat{\pi}_0(1;x), \ldots, \hat{\pi}_0(K;x)), where π^0(k;x)\hat{\pi}_0(k;x) represents the base estimate of p(k;x)p(k;x). Denote the feature embedding of xx by z:=ϕ(x)z := \phi(x) and write zi=ϕ(xi)z_i = \phi(x_i), i∈[n]:={1,…,n}i \in[n] := \{1,\ldots,n\}. The initial goal of LAPE is to construct a calibrated estimator of p(k;x)p(k;x), denoted by p^(k;x)\hat{p}(k;x), by leveraging π^0(k;x)\hat{\pi}_0(k;x), Dcal\mathcal{D}_{\mathrm{cal}}, and {zi}i=1n\{z_i\}_{i=1}^{n}.

Furthermore, let fcal(⋅;π^0,D):X→ΔK−1f_{\mathrm{cal}}(\cdot;\hat{\pi}_0,\mathcal{D}) : \mathcal{X} \to\Delta^{K-1} be any existing post-hoc calibration algorithm, which calibrates the initial soft-classifier π^0\hat{\pi}_0 based on a calibration dataset D⊆Dcal\mathcal{D} \subseteq\mathcal{D}_{\mathrm{cal}}. For example, the mapping fcalf_{\mathrm{cal}} can be defined by calibration algorithms such as VS and TS. We aim to utilize LAPE to further improve these calibration methods.

Motivation

A learned representation can be discriminative for classification while still leaving calibration errors that vary across different feature-space neighborhoods. For example, global post-hoc calibrators typically learn a fixed transformation from logits or confidence scores to calibrated probabilities. As a result, two samples with similar scores receive similar corrections, even if their nearby calibration samples exhibit different empirical label frequencies [38, 27, 10]. Such “same-score, different-reliability” implies a possible route beyond purely score-based calibration: retain the global calibrator as a stable anchor and use nearby held-out samples as local evidence for residual mismatch, when the feature geometry makes such neighborhoods informative.

The key insight of LAPE is to use local frequencies in calibration samples as residual evidence for probability calibration. Figure 1 illustrates this mechanism in a controlled two-dimensional simulation, where training and calibration samples are generated from the same distribution, and only the training samples are used to fit the classifier. The figure highlights two observations as follows.

Representation separation in raw and learned feature spaces

Figure 1. Representation separation can not guarantee uniform calibration support. (a) Raw features of training and calibration samples. (b) Learned features separate the training samples clearly, but calibration samples can remain mixed near induced boundary regions. Different colors represent different classes. The highlighted neighborhood illustrates the local evidence used by LAPE.

Observation 1. After representation learning, the training samples become more clearly separated, whereas held-out calibration samples can still remain locally mixed near induced boundary regions.

Observation 2. For held-out points with class-1 softmax scores in [0.35,0.65][0.35, 0.65], local frequencies from 32 nearest neighbors in the full learned feature space better approximate oracle probabilities than softmax predictions (MAE: 0.126 vs. 0.156).

Observation 1 reflects that the classifier is optimized to separate the observed training dataset but can fail to separate held-out samples near classification boundaries. Observation 2 demonstrates that nearby held-out calibration samples can provide empirical evidence about the local class composition that is not directly used by a purely score-based global calibrator. LAPE therefore uses this local frequency as a residual correction to a global calibrated anchor.

Evidence beyond two dimensions. We further examine the above observations in a controlled simulation with 100-dimensional inputs. We generate i.i.d. training, calibration, and test sets, each containing 2,000 samples. The classifier is obtained on the training dataset. For a split A∈{train,cal}A \in\{\mathrm{train}, \mathrm{cal}\} and a representation h∈{raw,learned}h \in\{\mathrm{raw}, \mathrm{learned}\}, let μA,ch\mu^{h}_{A,c} denote the centroid of class cc:

μA,ch=1∣IA,c∣∑i∈IA,ch(xi),IA,c={i:xi∈A, yi=c}.\mu^{h}_{A,c} = \frac{1}{\lvert\mathcal{I}_{A,c}\rvert}\sum_{i\in\mathcal{I}_{A,c}} h(x_i), \qquad\mathcal{I}_{A,c} = \{i : x_i \in A,\, y_i = c\}.

We use SAh=∥μA,1h−μA,0h∥2S^{h}_{A} = \lVert\mu^{h}_{A,1} - \mu^{h}_{A,0} \rVert_{2} as a class-separation diagnostic, after standardizing each representation using statistics from the training split.

Results are averaged over ten random seeds. In the raw input space, the training and calibration sets have nearly identical class separation (2.510 ± 0.176 versus 2.519 ± 0.162). After representation learning, the separation is moderately larger for the training samples than for the held-out calibration samples (3.663 ± 0.343 versus 2.981 ± 0.234), showing that a train–calibration discrepancy can remain even under i.i.d. sampling.

We then evaluate probability estimation on an independent test set. The boundary region is specified before model fitting as 0.35≤P(Y=1∣X)≤0.650.35 \le P(Y = 1 \mid X) \le0.65. For each test point, the local estimate is computed from the Laplace-smoothed label frequency of its 50 nearest calibration samples in the learned representation. Within this predefined region, the mean absolute error to the oracle conditional probability is 0.1067±0.00980.1067 \pm0.0098 for the softmax estimate and 0.0969±0.00550.0969 \pm0.0055 for the local-frequency estimate. Over the entire test samples which includes points that fall outside the boundary region, however, softmax remains more accurate than the local estimate (0.0941 ±\pm 0.0054 versus 0.1058 ±\pm 0.0042). These results support using local calibration frequencies as a residual correction in selected regions, rather than as a replacement for the global prediction.

Local Adaptive Probability Estimation

This section presents LAPE in two parts. We first introduce the general LAPE method, which constructs a non-parametric residual correction for a pre-trained soft-classifier π^0\hat{\pi}_0. We then describe how to apply LAPE to existing post-hoc calibrators by treating their calibrated outputs as π^0\hat{\pi}_0 under the shared calibration dataset Dcal\mathcal{D}_{\mathrm{cal}}.

General Methodology

We introduce the general method of LAPE in three steps. We first describe how to identify the neighbors of a given test point. Then, we introduce the hybrid refinement, which fuses the global anchor with this local evidence to produce the final calibrated probability. Finally, we describe how to employ the adaptive optimization to solve for the correction parameters that minimize residual errors.

Step 1: Local evidence retrieval. Given a new test sample xx and its feature embedding vector zz, the first step of LAPE is to select samples from Dcal\mathcal{D}_{\mathrm{cal}} that are geometrically close to xx in the feature space. Specifically, let di=∥zi−z∥2d_i = \lVert z_i-z\rVert_2, i∈[n]i \in[n]. For any fixed m∈[n]m \in[n], the set of mm nearest neighbors Nm(x)\mathcal{N}_m(x) is defined by

Nm(x):={i∈[n]:∥z−zi∥2≤d(m)},\mathcal{N}_m(x) := \{i \in[n] : \lVert z-z_i\rVert_2 \le d_{(m)}\},

where d(m)d_{(m)} denotes the mm-th smallest value of {di}i=1n\{d_i\}_{i=1}^n. By leveraging the local calibration dataset Nm(x)\mathcal{N}_m(x), LAPE then computes the local empirical distribution as

p~0(k;x,m):=∑i∈Nm(x)I(yi=k)+1/Km+1,k∈[K],(1)\tilde{p}_0(k;x,m) := \frac{\sum_{i\in\mathcal{N}_m(x)} \mathbb{I}(y_i=k)+1/K}{m+1}, \qquad k\in[K], \tag*{(1)}

where 1/K1/K is a smoothing term to avoid zero probabilities.

Step 2: Hybrid refinement. Let πanchor(k;x)\pi^{\mathrm{anchor}}(k;x) be the anchor estimate of p(k;x)p(k;x), k∈[K]k \in[K]. We focus on the case that

πanchor(k;x)=π^0(k;x).\pi^{\mathrm{anchor}}(k;x)=\hat{\pi}_0(k;x).

The idea of hybrid refinement is that, if the base prediction π^0(⋅)\hat{\pi}_0(\cdot) is already reliable in many regions, then it is unnecessary to adopt a large local correction. For sparse regions where π^0(⋅)\hat{\pi}_0(\cdot) performs not satisfactorily, local frequencies may also be weakly supported, which means that over-reliance on the local evidence can further increase variance. Therefore, LAPE treats p~0(k;x,m)\tilde{p}_0(k;x,m) as a delicate residual correction signal rather than a replacement for the global anchor. Specifically, given a neighborhood size m∈[n]m \in[n] and a weight α∈[0,1]\alpha\in[0,1], the calibrated probability of π^0(k;x)\hat{\pi}_0(k;x) is given by

p^(k;x,m,α):=(1−α)⋅πanchor(k;x)+α⋅p~0(k;x,m),k∈[K].(2)\hat{p}(k;x,m,\alpha) := (1-\alpha)\cdot\pi^{\mathrm{anchor}}(k;x)+\alpha\cdot\tilde{p}_0(k;x,m), \qquad k\in[K]. \tag*{(2)}

In what follows, we describe how to choose mm and α\alpha adaptively.

Step 3: Adaptive hyperparameter selection. LAPE selects the hyperparameter vector (m,α)(m,\alpha) by leveraging weighted leave-one-out (LOO) analysis. Specifically, for any fixed i∈[n]i \in[n], we denote the LOO probability estimate of xix_i by

p(i)(yi;xi,m,α):=(1−α)πanchor(yi;xi)+αp~−i(yi;xi,m).(3)p^{(i)}(y_i;x_i,m,\alpha):=(1-\alpha)\pi^{\mathrm{anchor}}(y_i;x_i)+\alpha\tilde{p}_{-i}(y_i;x_i,m). \tag*{(3)}

where p~−i(k;xi,m)\tilde{p}_{-i}(k;x_i,m) is the LOO local empirical distribution that

p~−i(k;xi,m):=∑j∈Nm(x)∖{i}I(yj=k)+1/Km,k∈[K].(4)\tilde{p}_{-i}(k;x_i,m):=\frac{\sum_{j\in\mathcal{N}_m(x)\setminus\{i\}}\mathbb{I}(y_j=k)+1/K}{m}, \qquad k\in[K]. \tag*{(4)}

Each calibration sample is weighted by its distance to the feature embedding of the test point xx, i.e.,

wi(m)=exp⁡(−∥z−zi∥22/σ2)∑j∈Nm(x)exp⁡(−∥z−zj∥22/σ2).(5)w_i^{(m)}=\frac{\exp\left(-\lVert z-z_i\rVert_2^2/\sigma^2\right)}{\sum_{j\in\mathcal{N}_m(x)}\exp\left(-\lVert z-z_j\rVert_2^2/\sigma^2\right)}. \tag*{(5)}

References

  1. [1]Ethem Alpaydin and Cenk Kaynak. Optical recognition of handwritten digits. UCI Machine Learning Repository, 1998.
  2. [2]Dario Amodei, Chris Olah, Jacob Steinhardt, Paul Christiano, John Schulman, and Dan Mané. Concrete problems in ai safety. arXiv preprint arXiv:1606.06565, 2016.arxiv.org/abs/1606.06565
  3. [3]Sergio A. Balanya, Juan Maroñas, and Daniel Ramos. Adaptive temperature scaling for robust calibration of deep neural networks. Neural Computing and Applications, 2024.
  4. [4]Tadas Baltrušaitis, Chaitanya Ahuja, and Louis-Philippe Morency. Multimodal machine learning: A survey and taxonomy. IEEE Transactions on Pattern Analysis and Machine Intelligence, 41(2): 423–443, 2019.DOI
  5. [5]Jon Louis Bentley. Multidimensional binary search trees used for associative searching. Communications of the ACM, 18(9):509–517, 1975.DOI
  6. [6]Jock A Blackard and Denis J Dean. Comparative accuracies of artificial neural networks and discriminant analysis in predicting forest cover types from cartographic variables. Computers and Electronics in Agriculture, 24(3):131–151, 1999.DOI
  7. [7]Bing Cao, Yinan Xia, Yi Ding, Changqing Zhang, and Qinghua Hu. Predictive dynamic fusion. In Proceedings of the 41st International Conference on Machine Learning, volume 235 of Proceedings of Machine Learning Research, pages 5608–5628. PMLR, 2024.arxiv.org/abs/2406.04802
  8. [8]Mircea Cimpoi, Subhransu Maji, Iasonas Kokkinos, Sammy Mohamed, and Andrea Vedaldi. Describing textures in the wild. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 3606–3613, 2014. doi: 10.1109/CVPR.2014.461.DOI
  9. [9]Frances Ding, Moritz Hardt, John Miller, and Ludwig Schmidt. Retiring adult: New datasets for fair machine learning. Advances in Neural Information Processing Systems, 34:6478–6490, 2021a.arxiv.org/abs/2108.04884
  10. [10]Zhipeng Ding, Xu Han, Peirong Liu, and Marc Niethammer. Local temperature scaling for probability calibration. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 6889–6899, October 2021b.arxiv.org/abs/2008.05105
  11. [11]Robert Duin. Multiple Features. UCI Machine Learning Repository, 1998.DOI
  12. [12]Peter W Frey and David J Slate. Letter recognition using holland-style adaptive classifiers. Machine Learning, 6(2):161–182, 1991.
  13. [13]Tilmann Gneiting and Adrian E Raftery. Strictly proper scoring rules, prediction, and estimation. Journal of the American Statistical Association, 102(477):359–378, 2007.
  14. [14]Chuan Guo, Geoff Pleiss, Yu Sun, and Kilian Q Weinberger. On calibration of modern neural networks. In International Conference on Machine Learning, pages 1321–1330. PMLR, 2017.
  15. [15]Ramya Hebbalaguppe, Jatin Prakash, Neelabh Madan, and Chetan Arora. A stitch in time saves nine: A train-time regularizing loss for improved neural network calibration. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 16081–16090, 2022.arxiv.org/abs/2203.13834
  16. [16]Dan Hendrycks and Thomas Dietterich. Benchmarking neural network robustness to common corruptions and perturbations. In International Conference on Learning Representations, 2019.
  17. [17]Heinrich Jiang, Been Kim, Melody Guan, and Maya Gupta. To trust or not to trust a classifier. Advances in Neural Information Processing Systems, 31, 2018.
  18. [18]Tom Joy, Francesco Pinto, Ser-Nam Lim, Philip H. S. Torr, and Puneet K. Dokania. Sample-dependent adaptive temperature scaling for improved calibration. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 14919–14926, 2023.arxiv.org/abs/2207.06211
  19. [19]Ross D. King, Cao Feng, and Alistair Sutherland. Statlog: comparison of classification algorithms on large real-world problems. Applied Artificial Intelligence an International Journal, 9(3):289–333, 1995.DOI
  20. [20]Ron Kohavi et al. Scaling up the accuracy of naive-bayes classifiers: A decision-tree hybrid. In Proceedings of the Second International Conference on Knowledge Discovery and Data Mining, volume 96, pages 202–207, 1996.
  21. [21]Alex Krizhevsky and Geoffrey Hinton. Learning multiple layers of features from tiny images. Technical report, University of Toronto, Toronto, Ontario, 2009.
  22. [22]Meelis Kull, Telmo M Silva Filho, and Peter Flach. Beyond sigmoids: How to obtain well-calibrated probabilities from binary classifiers with beta calibration. Electronic Journal of Statistics, 11(2):5052–5080, 2017.DOI
  23. [23]Meelis Kull, Miquel Perello Nieto, Markus Kängsepp, Telmo Silva Filho, Hao Song, and Peter Flach. Beyond temperature scaling: Obtaining well-calibrated multi-class probabilities with dirichlet calibration. Advances in Neural Information Processing Systems, 32, 2019.
  24. [24]Ananya Kumar, Percy S. Liang, and Tengyu Ma. Verified uncertainty calibration. In Advances in Neural Information Processing Systems, volume 32, 2019.arxiv.org/abs/1909.10155
  25. [25]Balaji Lakshminarayanan, Alexander Pritzel, and Charles Blundell. Simple and scalable predictive uncertainty estimation using deep ensembles. Advances in Neural Information Processing Systems, 30, 2017.arxiv.org/abs/1612.01474
  26. [26]Bingyuan Liu, Ismail Ben Ayed, Adrian Galdran, and Jose Dolz. The devil is in the margin: Margin-based label smoothing for network calibration. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 80–88, 2022.arxiv.org/abs/2111.15430
  27. [27]Rachel Luo, Aadyot Bhatnagar, Yu Bai, Shengjia Zhao, Huan Wang, Caiming Xiong, Silvio Savarese, Stefano Ermon, Edward Schmerling, and Marco Pavone. Local calibration: Metrics and recalibration. In James Cussens and Kun Zhang, editors, Proceedings of the Thirty-Eighth Conference on Uncertainty in Artificial Intelligence, volume 180 of Proceedings of Machine Learning Research, pages 1286–1295. PMLR, 2022.arxiv.org/abs/2102.10809
  28. [28]Pete Mowforth and Barry Shepherd. Statlog (Vehicle Silhouettes) [dataset]. UCI Machine Learning Repository, 1987.DOI
  29. [29]Mahdi Pakdaman Naeini, Gregory Cooper, and Milos Hauskrecht. Obtaining well calibrated probabilities using bayesian binning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 29, 2015.pubmed.ncbi.nlm.nih.gov/25927013
  30. [30]Jeremy Nixon, Michael W. Dusenberry, Linchuan Zhang, Ghassen Jerfel, and Dustin Tran. Measuring calibration in deep learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2019.
  31. [31]Stephen M. Omohundro. Five balltree construction algorithms. Technical Report TR-89-063, International Computer Science Institute, Berkeley, CA, 1989.
  32. [32]Fabio Mendoza Palechor and Alexis De la Hoz Manotas. Dataset for estimation of obesity levels based on eating habits and physical condition in individuals from colombia, peru and mexico. Data in Brief, 25:104344, 2019.DOI
  33. [33]Andrea Dal Pozzolo, Olivier Caelen, Reid A. Johnson, and Gianluca Bontempi. Calibrating probability with undersampling for unbalanced classification. In 2015 IEEE Symposium Series on Computational Intelligence, pages 159–166, 2015.
  34. [34]Byron P Roe, Hai-Jun Yang, Ji Zhu, Yong Liu, Ion Stancu, and Gordon McGregor. Boosted decision trees as an alternative to artificial neural networks for particle identification. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 543(2-3):577–584, 2005.DOI
  35. [35]Oliver Roesler. EEG Eye State [dataset]. UCI Machine Learning Repository, 2013.DOI
  36. [36]Cees G. M. Snoek, Marcel Worring, and Arnold W. M. Smeulders. Early versus late fusion in semantic video analysis. In Proceedings of the 13th Annual ACM International Conference on Multimedia, pages 399–402. Association for Computing Machinery, 2005.DOI
  37. [37]Linwei Tao, Minjing Dong, and Chang Xu. Feature clipping for uncertainty calibration. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 20841–20849, 2025.arxiv.org/abs/2410.19796
  38. [38]Christian Tomani, Daniel Cremers, and Florian Buettner. Parameterized temperature scaling for boosting the expressive power in post-hoc uncertainty calibration. In European Conference on Computer Vision, pages 555–569. Springer, 2022.arxiv.org/abs/2102.12182
  39. [39]Philipp Tschandl, Cliff Rosendahl, and Harald Kittler. The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions. Scientific Data, 5:180161, 2018.
  40. [40]UCI Machine Learning Repository. Image Segmentation [dataset]. UCI Machine Learning Repository, 1990.
  41. [41]Shuoyuan Wang, Jindong Wang, Guoqing Wang, Bob Zhang, Kaiyang Zhou, and Hongxin Wei. Open-vocabulary calibration for fine-tuned CLIP. In Proceedings of the 41st International Conference on Machine Learning, volume 235 of Proceedings of Machine Learning Research, pages 51734–51754. PMLR, 2024.arxiv.org/abs/2402.04655
  42. [42]Shuoyuan Wang, Beier Luo, Hao Zeng, Chengyao Yu, Songxin Zhang, Zejian Xie, Bingyi Jing, and Hongxin Wei. On the pitfalls of verbalized confidence priors for calibrating large reasoning models. arXiv preprint arXiv:2609.32470, 2026.arxiv.org/abs/2609.32470
  43. [43]Jerry Wei, Arief Suriawinata, Bing Ren, Xiaoying Liu, Mikhail Lisovsky, Louis Vaickus, Charles Brown, Michael Baker, Naofumi Tomita, Lorenzo Torresani, et al. A petri dish for histopathology image analysis. In International Conference on Artificial Intelligence in Medicine, pages 11–24. Springer, 2021.arxiv.org/abs/2101.12355
  44. [44]David H. Wolpert. Stacked generalization. Neural Networks, 5(2):241–259, 1992.DOI
  45. [45]Chang Xu, Dacheng Tao, and Chao Xu. A survey on multi-view learning. arXiv preprint arXiv:1304.5634, 2013.arxiv.org/abs/1304.5634
  46. [46]Jia-Qi Yang, De-Chuan Zhan, and Le Gan. Beyond probability partitions: Calibrating neural networks with semantic aware grouping. In Advances in Neural Information Processing Systems, volume 36, pages 58448–58460, 2023a.arxiv.org/abs/2306.04985
  47. [47]Jiancheng Yang, Rui Shi, Donglai Wei, Zequan Liu, Lin Zhao, Bilian Ke, Hanspeter Pfister, and Bingbing Ni. Medmnist v2 - a large-scale lightweight benchmark for 2d and 3d biomedical image classification. Scientific Data, 10(1):41, 2023b.
  48. [48]Chengyao Yu, Hongxin Wei, and Bingyi Jing. Robust conformalized selection with noisy responses. arXiv preprint arXiv:2607.22985, 2026a.arxiv.org/abs/2607.22985
  49. [50]Bianca Zadrozny and Charles Elkan. Obtaining calibrated probability estimates from decision trees and naive bayesian classifiers. In International Conference on Machine Learning, volume 1, 2001.
  50. [51]Bianca Zadrozny and Charles Elkan. Transforming classifier scores into accurate multiclass probability estimates. In Proceedings of the eighth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 694–699, 2002.
  51. [52]Jize Zhang, Bhavya Kalluri, and T Yong-Jin Han. Mix-n-match: Ensemble and compositional methods for uncertainty calibration in deep learning. In International Conference on Machine Learning, pages 11117–11128. PMLR, 2020.
  52. [53]Qingyang Zhang, Haitao Wu, Changqing Zhang, Qinghua Hu, Huazhu Fu, Joey Tianyi Zhou, and Xi Peng. Provable dynamic fusion for low-quality multimodal data. In Proceedings of the 40th International Conference on Machine Learning, volume 202 of Proceedings of Machine Learning Research, pages 41753–41769. PMLR, 2023.arxiv.org/abs/2306.02050
  53. [54]Kaiyang Zhou, Jingkang Yang, Chen Change Loy, and Ziwei Liu. Learning to prompt for vision-language models. International Journal of Computer Vision, 130(9):2337–2348, 2022.DOI

Paper details

vlnc:W_j8xakj6u
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Computer Science

Not recorded for this paper

View statement
Looking for related papers…
Contents