Arrow Research search

Author name cluster

Lei Feng

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

48 papers
2 author rows

Possible papers

48

AAAI Conference 2026 Conference Paper

An Invariant Latent Space Perspective on Language Model Inversion

  • Wentao Ye
  • Jiaqi Hu
  • Haobo Wang
  • Xinpeng Ti
  • Zhiqing Xiao
  • Hao Chen
  • Liyao Li
  • Lei Feng

Language model inversion (LMI), i.e., recovering hidden prompts from outputs, emerges as a concrete threat to user privacy and system security. We recast LMI as reusing the LLM's own latent space and propose the Invariant Latent Space Hypothesis (ILSH): (1) diverse outputs from the same source prompt should preserve consistent semantics (source invariance), and (2) input output cyclic mappings should be self-consistent within a shared latent space (cyclic invariance). Accordingly, we present Inv2A, which treats the LLM as an invariant decoder and learns only a lightweight inverse encoder that maps outputs to a denoised pseudo-representation. When multiple outputs are available, they are sparsely concatenated at the representation layer to increase information density. Training proceeds in two stages: contrastive alignment (source invariance) and supervised reinforcement (cyclic invariance). An optional training-free neighborhood search can refine local performance. Across 9 datasets covering user and system prompt scenarios, Inv2A outperforms baselines by an average of 4.77% BLEU score while reducing dependence on large inverse corpora. Our analysis further shows that prevalent defenses provide limited protection, underscoring the need for stronger strategies.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

Defending Multimodal Backdoored Models by Repulsive Visual Prompt Tuning

  • Zhifang Zhang
  • Shuo He
  • Haobo Wang
  • Bingquan Shen
  • Lei Feng

Multimodal contrastive learning models (e. g. , CLIP) can learn high-quality representations from large-scale image-text datasets, while they exhibit significant vulnerabilities to backdoor attacks, raising serious safety concerns. In this paper, we reveal that CLIP's vulnerabilities primarily stem from its tendency to encode features beyond in-dataset predictive patterns, compromising its visual feature resistivity to input perturbations. This makes its encoded features highly susceptible to being reshaped by backdoor triggers. To address this challenge, we propose Repulsive Visual Prompt Tuning (RVPT), a novel defense approach that employs deep visual prompt tuning with a specially designed feature-repelling loss. Specifically, RVPT adversarially repels the encoded features from deeper layers while optimizing the standard cross-entropy loss, ensuring that only predictive features in downstream tasks are encoded, thereby enhancing CLIP’s visual feature resistivity against input perturbations and mitigating its susceptibility to backdoor attacks. Unlike existing multimodal backdoor defense methods that typically require the availability of poisoned data or involve fine-tuning the entire model, RVPT leverages few-shot downstream clean samples and only tunes a small number of parameters. Empirical results demonstrate that RVPT tunes only 0. 27\% of the parameters in CLIP, yet it significantly outperforms state-of-the-art defense methods, reducing the attack success rate from 89. 70\% to 2. 76\% against the most advanced multimodal attacks on ImageNet and effectively generalizes its defensive capabilities across multiple datasets. Our code is available on https: //anonymous. 4open. science/r/rvpt-anonymous.

PDF Details

TMLR Journal 2025 Journal Article

Does confidence calibration improve conformal prediction?

  • HuaJun Xi
  • Jianguo Huang
  • Kangdao Liu
  • Lei Feng
  • Hongxin Wei

Conformal prediction is an emerging technique for uncertainty quantification that constructs prediction sets guaranteed to contain the true label with a predefined probability. Previous works often employ temperature scaling to calibrate classifiers, assuming that confidence calibration benefits conformal prediction. However, the specific impact of confidence calibration on conformal prediction remains underexplored. In this work, we make two key discoveries about the impact of confidence calibration methods on adaptive conformal prediction. Firstly, we empirically show that current confidence calibration methods (e.g., temperature scaling) typically lead to larger prediction sets in adaptive conformal prediction. Secondly, by investigating the role of temperature value, we observe that high-confidence predictions can enhance the efficiency of adaptive conformal prediction. Theoretically, we prove that predictions with higher confidence result in smaller prediction sets on expectation. This finding implies that the rescaling parameters in these calibration methods, when optimized with cross-entropy loss, might counteract the goal of generating efficient prediction sets. To address this issue, we propose \textbf{Conformal Temperature Scaling} (ConfTS), a variant of temperature scaling with a novel loss function designed to enhance the efficiency of prediction sets. This approach can be extended to optimize the parameters of other post-hoc methods of confidence calibration. Extensive experiments demonstrate that our method improves existing adaptive conformal prediction methods in both image and text classification tasks.

PDF Details

NeurIPS Conference 2025 Conference Paper

Enhancing Sample Selection Against Label Noise by Cutting Mislabeled Easy Examples

  • Suqin Yuan
  • Lei Feng
  • Bo Han
  • Tongliang Liu

Sample selection is a prevalent approach in learning with noisy labels, aiming to identify confident samples for training. Although existing sample selection methods have achieved decent results by reducing the noise rate of the selected subset, they often overlook that not all mislabeled examples harm the model's performance equally. In this paper, we demonstrate that mislabeled examples correctly predicted by the model early in the training process are particularly harmful to model performance. We refer to these examples as Mislabeled Easy Examples (MEEs). To address this, we propose Early Cutting, which introduces a recalibration step that employs the model's later training state to re-select the confident subset identified early in training, thereby avoiding misleading confidence from early learning and effectively filtering out MEEs. Experiments on the CIFAR, WebVision, and full ImageNet-1k datasets demonstrate that our method effectively improves sample selection and model performance by reducing MEEs.

PDF Details

NeurIPS Conference 2025 Conference Paper

Establishing Linear Surrogate Regret Bounds for Convex Smooth Losses via Convolutional Fenchel–Young Losses

  • Yuzhou Cao
  • Han Bao
  • Lei Feng
  • Bo An

Surrogate regret bounds, also known as excess risk bounds, bridge the gap between the convergence rates of surrogate and target losses. The regret transfer is lossless if the surrogate regret bound is linear. While convex smooth surrogate losses are appealing in particular due to the efficient estimation and optimization, the existence of a trade-off between the loss smoothness and linear regret bound has been believed in the community. Under this scenario, the better optimization and estimation properties of convex smooth surrogate losses may inevitably deteriorate after undergoing the regret transfer onto a target loss. We overcome this dilemma for arbitrary discrete target losses by constructing a convex smooth surrogate loss, which entails a linear surrogate regret bound composed with a tailored prediction link. The construction is based on Fenchel--Young losses generated by the convolutional negentropy, which are equivalent to the infimal convolution of a generalized negentropy and the target Bayes risk. Consequently, the infimal convolution enables us to derive a smooth loss while maintaining the surrogate regret bound linear. We additionally benefit from the infimal convolution to have a consistent estimator of the underlying class probability. Our results are overall a novel demonstration of how convex analysis penetrates into optimization and statistical efficiency in risk minimization.

PDF Details

TMLR Journal 2025 Journal Article

Exploring Weak-to-Strong Generalization for CLIP-based Classification

  • Jinhao Li
  • Sarah Monazam Erfani
  • Lei Feng
  • James Bailey
  • Feng Liu

Aligning large-scale commercial models with user intent is crucial to preventing harmful outputs. Current methods rely on human supervision but become impractical as model complexity increases. When models surpass human knowledge, providing accurate feedback becomes challenging and inefficient. A novel solution proposed recently is using a weaker model to supervise a stronger model. This concept leverages the ability of weaker models to perform evaluations, thereby reducing the workload on human supervisors. Previous work has shown the effectiveness of weak-to-strong generalization in the context of language-only models. Extending this concept to vision-language models leverages these insights, adapting the proven benefits to a multi-modal context. In our study, we explore weak-to-strong generalization for CLIP-based classification. We propose a method, \emph{class prototype learning} (CPL), which aims to enhance the classification capabilities of the CLIP model, by learning more representative prototypes for each category. Our findings indicate that, despite using a simple loss function under weak supervision, CPL yields robust improvements in targeted scenarios, particularly when pretraining is limited. Extensive experiments demonstrate that our approach is effective under these settings, achieving a 3.67\% improvement over strong baseline methods.

PDF Details

AAAI Conference 2025 Conference Paper

Improving Generalization of Deep Neural Networks by Optimum Shifting

  • Yuyan Zhou
  • Ye Li
  • Lei Feng
  • Sheng-Jun Huang

Recent studies showed that the generalization of neural networks is correlated with the sharpness of the loss landscape and flat minima suggests a better generalization ability than sharp minima. In this paper, we propose a novel method called optimum shifting, which changes the parameters of a neural network from a sharp minimum to a flatter one while maintaining the same training loss value. Our method is based on the observation that when the input and output of a neural network are fixed, the matrix multiplications within the network can be treated as systems of under-determined linear equations, enabling adjustment of parameters in the solution space, which can be simply accomplished by solving a constrained optimization problem. Furthermore, we introduce a practical stochastic optimum shifting technique utilizing the neural collapse theory to reduce computational costs and provide more degrees of freedom for optimum shifting. Extensive experiments with various deep neural network architectures on benchmark datasets demonstrate the effectiveness of our method.

PDF Details DOI

AAAI Conference 2025 Conference Paper

Influence-Based Fair Selection for Sample-Discriminative Backdoor Attack

  • Qi Wei
  • Shuo He
  • Jiahan Zhang
  • Lei Feng
  • Bo An

Backdoor attacks have posed a serious threat in machine learning models, wherein adversaries can poison training samples with maliciously crafted triggers to compromise the victim model. Advanced backdoor attack methods have focused on selectively poisoning more vulnerable training samples, achieving a higher attack success rate (ASR). However, we found that when the manipulation strength of the trigger is constrained to a very small value for imperceptible attacks, they suffer from extremely uneven class-wise ASR due to the unequal selection of instances per class. To solve this issue, we propose a novel backdoor attack method based on Influence-based Fair Selection (IFS), including two objectives: 1) selecting samples that significantly contribute to ASR and 2) ensuring class balance during the selection process. Specifically, we adapt Influence Functions, a classic technique in robust statistics, to evaluate the influence of trigger-embedded training samples on ASR. In this case, training samples contributing to reducing the backdoored test risk could possess higher influence scores. Further, a group-based pruning strategy is designed to avoid calculating the influence on ASR for all training samples, thereby significantly reducing the computational cost. Then, based on the influence score, we design an adaptive thresholding scheme to dynamically select samples with higher influence while maintaining class balance. Extensive experiments on four datasets verify the effectiveness of IFS compared with advanced methods.

PDF Details DOI

IJCAI Conference 2025 Conference Paper

Prototype-based Optimal Transport for Out-of-Distribution Detection

  • Ao Ke
  • Wenlong Chen
  • Chuanwen Feng
  • Yukun Cao
  • Xike Xie
  • S. Kevin Zhou
  • Lei Feng

Detecting Out-of-Distribution (OOD) inputs is crucial for improving the reliability of deep neural networks in the real-world deployment. In this paper, inspired by the inherent distribution shift between in-distribution (ID) and OOD data, we propose a novel method that leverages optimal transport to measure the distribution discrepancy between test inputs and ID prototypes. The resulting transport costs are used to quantify the individual contribution of each test input to the overall discrepancy, serving as a desirable measure for OOD detection. To address the issue that solely relying on the transport costs to ID prototypes is inadequate for identifying OOD inputs closer to ID data, we generate virtual outliers to approximate the OOD region via linear extrapolation. By combining the transport costs to ID prototypes with the costs to virtual outliers, the detection of OOD data near ID data is emphasized, thereby enhancing the distinction between ID and OOD inputs. Extensive evaluations demonstrate the superiority of our method over state-of-the-art methods.

PDF Details DOI

JBHI Journal 2025 Journal Article

rU-Net, Multi-Scale Feature Fusion and Transfer Learning: Unlocking the Potential of Cuffless Blood Pressure Monitoring With PPG and ECG

  • Jiaming Chen
  • Xueling Zhou
  • Lei Feng
  • Bingo Wing-Kuen Ling
  • Lianyi Han
  • Hongtao Zhang

This study introduces an innovative deep-learning model for cuffless blood pressure estimation using PPG and ECG signals, demonstrating state-of-the-art performance on the largest clean dataset, PulseDB. The rU-Net architecture, a fusion of U-Net and ResNet, enhances both generalization and feature extraction accuracy. Accurate multi-scale feature capture is facilitated by short-time Fourier transform (STFT) time-frequency distributions and multi-head attention mechanisms, allowing data-driven feature selection. The inclusion of demographic parameters as supervisory information further elevates performance. On the calibration-based dataset, our model excels, achieving outstanding accuracy (SBP MAE ± std: 4. 49 ± 4. 86 mmHg, DBP MAE ± std: 2. 69 ± 3. 10 mmHg), surpassing AAMI standards and earning a BHS Grade A rating. Addressing the challenge of calibration-free data, we propose a fine-tuning-based transfer learning approach. Remarkably, with only 10% data transfer, our model attains exceptional accuracy (SBP MAE ± std: 4. 14 ± 5. 01 mmHg, DBP MAE ± std: 2. 48 ± 2. 93 mmHg). This study sets the stage for the development of highly accurate and reliable wearable cuffless blood pressure monitoring devices.

Details DOI

IJCAI Conference 2025 Conference Paper

Towards Robust Incremental Learning Under Ambiguous Supervision

  • Rui Wang
  • Mingxuan Xia
  • Haobo Wang
  • Lei Feng
  • Junbo Zhao
  • Gang Chen
  • Chang Yao

Traditional Incremental Learning (IL) targets to handle sequential fully-supervised learning problems where novel classes emerge from time to time. However, due to inherent annotation uncertainty and ambiguity, collecting high-quality annotated data in a dynamic learning system can be extremely expensive. To mitigate this problem, we propose a novel weakly-supervised learning paradigm called Incremental Partial Label Learning (IPLL), where the sequentially arrived data relate to a set of candidate labels rather than the ground truth. Technically, we develop the Prototype-Guided Disambiguation and Replay Algorithm (PGDR) which leverages the class prototypes as a proxy to mitigate two intertwined challenges in IPLL, i. e. , label ambiguity and catastrophic forgetting. To handle the former, PGDR encapsulates a momentum-based pseudo-labeling algorithm along with prototype-guided initialization, resulting in a balanced perception of classes. To alleviate forgetting, we develop a memory replay technique that collects well-disambiguated samples while maintaining representativeness and diversity. By jointly distilling knowledge from curated memory data, our framework exhibits a great disambiguation ability for samples of new tasks and achieves less forgetting of knowledge. Extensive experiments demonstrate that PGDR achieves superior performance over the baselines in the IPLL task.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

Bayesian-guided Label Mapping for Visual Reprogramming

  • Chengyi Cai
  • Zesheng Ye
  • Lei Feng
  • Jianzhong Qi
  • Feng Liu

Visual reprogramming (VR) leverages the intrinsic capabilities of pretrained vision models by adapting their input or output interfaces to solve downstream tasks whose labels (i. e. , downstream labels) might be totally different from the labels associated with the pretrained models (i. e. , pretrained labels). When adapting the output interface, label mapping methods transform the pretrained labels to downstream labels by establishing a gradient-free one-to-one correspondence between the two sets of labels. However, in this paper, we reveal that one-to-one mappings may overlook the complex relationship between pretrained and downstream labels. Motivated by this observation, we propose a ** B ayesian-guided L abel M apping (BLM) method. BLM constructs an iteratively-updated probabilistic label mapping matrix, with each element quantifying a pairwise relationship between pretrained and downstream labels. The assignment of values to the constructed matrix is guided by Bayesian conditional probability, considering the joint distribution of the downstream labels and the labels predicted by the pretrained model on downstream samples. Experiments conducted on both pretrained vision models (e. g. , ResNeXt) and vision-language models (e. g. , CLIP) demonstrate the superior performance of BLM over existing label mapping methods. The success of BLM also offers a probabilistic lens through which to understand and analyze the effectiveness of VR. Our code is available at https: //github. com/tmlr-group/BayesianLM.

PDF Details DOI

ICRA Conference 2024 Conference Paper

Non-Axiomatic Reasoning for an Autonomous Mobile Robot

  • Patrick Hammer
  • Peter Isaev
  • Lei Feng
  • Robert Johansson
  • Jana Tumova

We present the integration of a Non-Axiomatic Reasoning System (NARS) with mobile robots for planning and decision making. NARS enables robots to effectively handle uncertainty in real-time with complete sensor and actuator integration, thereby ensuring adaptability to evolving scenarios. We discuss essential parts of the logic, the architecture and working principles of NARS, and the integration of NARS as a ROS node. A case study is provided demonstrating the system’s proficiency to carry out a garbage collection task in an open-air environment by operating a mobile robot with manipulator arm, and we demonstrate its ability to learn about the place-dependent accumulation of garbage items. Case study also reveals that our approach performs more effectively on the overall task than the Belief-Desire-Intention model we compared with.

Details

AAAI Conference 2024 Conference Paper

Robust Node Classification on Graph Data with Graph and Label Noise

  • Yonghua Zhu
  • Lei Feng
  • Zhenyun Deng
  • Yang Chen
  • Robert Amor
  • Michael Witbrock

Current research for node classification focuses on dealing with either graph noise or label noise, but few studies consider both of them. In this paper, we propose a new robust node classification method to simultaneously deal with graph noise and label noise. To do this, we design a graph contrastive loss to conduct local graph learning and employ self-attention to conduct global graph learning. They enable us to improve the expressiveness of node representation by using comprehensive information among nodes. We also utilize pseudo graphs and pseudo labels to deal with graph noise and label noise, respectively. Furthermore, We numerically validate the superiority of our method in terms of robust node classification compared with all comparison methods.

PDF Details DOI

AAAI Conference 2023 Conference Paper

A Generalized Unbiased Risk Estimator for Learning with Augmented Classes

  • Senlin Shu
  • Shuo He
  • Haobo Wang
  • Hongxin Wei
  • Tao Xiang
  • Lei Feng

In contrast to the standard learning paradigm where all classes can be observed in training data, learning with augmented classes (LAC) tackles the problem where augmented classes unobserved in the training data may emerge in the test phase. Previous research showed that given unlabeled data, an unbiased risk estimator (URE) can be derived, which can be minimized for LAC with theoretical guarantees. However, this URE is only restricted to the specific type of one-versus-rest loss functions for multi-class classification, making it not flexible enough when the loss needs to be changed with the dataset in practice. In this paper, we propose a generalized URE that can be equipped with arbitrary loss functions while maintaining the theoretical guarantees, given unlabeled data for LAC. To alleviate the issue of negative empirical risk commonly encountered by previous studies, we further propose a novel risk-penalty regularization term. Experiments demonstrate the effectiveness of our proposed method.

PDF Details DOI

NeurIPS Conference 2023 Conference Paper

ALIM: Adjusting Label Importance Mechanism for Noisy Partial Label Learning

  • Mingyu Xu
  • Zheng Lian
  • Lei Feng
  • Bin Liu
  • Jianhua Tao

Noisy partial label learning (noisy PLL) is an important branch of weakly supervised learning. Unlike PLL where the ground-truth label must conceal in the candidate label set, noisy PLL relaxes this constraint and allows the ground-truth label may not be in the candidate label set. To address this challenging problem, most of the existing works attempt to detect noisy samples and estimate the ground-truth label for each noisy sample. However, detection errors are unavoidable. These errors can accumulate during training and continuously affect model optimization. To this end, we propose a novel framework for noisy PLL with theoretical interpretations, called ``Adjusting Label Importance Mechanism (ALIM)''. It aims to reduce the negative impact of detection errors by trading off the initial candidate set and model outputs. ALIM is a plug-in strategy that can be integrated with existing PLL approaches. Experimental results on multiple benchmark datasets demonstrate that our method can achieve state-of-the-art performance on noisy PLL. Our code is available at: https: //github. com/zeroQiaoba/ALIM.

PDF Details

NeurIPS Conference 2023 Conference Paper

Binary Classification with Confidence Difference

  • Wei Wang
  • Lei Feng
  • Yuchen Jiang
  • Gang Niu
  • Min-Ling Zhang
  • Masashi Sugiyama

Recently, learning with soft labels has been shown to achieve better performance than learning with hard labels in terms of model generalization, calibration, and robustness. However, collecting pointwise labeling confidence for all training examples can be challenging and time-consuming in real-world scenarios. This paper delves into a novel weakly supervised binary classification problem called confidence-difference (ConfDiff) classification. Instead of pointwise labeling confidence, we are given only unlabeled data pairs with confidence difference that specifies the difference in the probabilities of being positive. We propose a risk-consistent approach to tackle this problem and show that the estimation error bound achieves the optimal convergence rate. We also introduce a risk correction approach to mitigate overfitting problems, whose consistency and convergence rate are also proven. Extensive experiments on benchmark data sets and a real-world recommender system data set validate the effectiveness of our proposed approaches in exploiting the supervision information of the confidence difference.

PDF Details

TIST Journal 2023 Journal Article

COMET: Convolutional Dimension Interaction for Collaborative Filtering

  • Zhuoyi Lin
  • Lei Feng
  • Xingzhi Guo
  • Yu Zhang
  • Rui Yin
  • Chee Keong Kwoh
  • Chi Xu

Representation learning-based recommendation models play a dominant role among recommendation techniques. However, most of the existing methods assume both historical interactions and embedding dimensions are independent of each other, and thus regrettably ignore the high-order interaction information among historical interactions and embedding dimensions. In this article, we propose a novel representation learning-based model called COMET ( CO nvolutional di M E nsion in T eraction), which simultaneously models the high-order interaction patterns among historical interactions and embedding dimensions. To be specific, COMET stacks the embeddings of historical interactions horizontally at first, which results in two “embedding maps”. In this way, internal interactions and dimensional interactions can be exploited by convolutional neural networks (CNN) with kernels of different sizes simultaneously. A fully connected multi-layer perceptron (MLP) is then applied to obtain two interaction vectors. Lastly, the representations of users and items are enriched by the learnt interaction vectors, which can further be used to produce the final prediction. Extensive experiments and ablation studies on various public implicit feedback datasets clearly demonstrate the effectiveness and rationality of our proposed method.

Details DOI

NeurIPS Conference 2023 Conference Paper

In Defense of Softmax Parametrization for Calibrated and Consistent Learning to Defer

  • Yuzhou Cao
  • Hussein Mozannar
  • Lei Feng
  • Hongxin Wei
  • Bo An

Enabling machine learning classifiers to defer their decision to a downstream expert when the expert is more accurate will ensure improved safety and performance. This objective can be achieved with the learning-to-defer framework which aims to jointly learn how to classify and how to defer to the expert. In recent studies, it has been theoretically shown that popular estimators for learning to defer parameterized with softmax provide unbounded estimates for the likelihood of deferring which makes them uncalibrated. However, it remains unknown whether this is due to the widely used softmax parameterization and if we can find a softmax-based estimator that is both statistically consistent and possesses a valid probability estimator. In this work, we first show that the cause of the miscalibrated and unbounded estimator in prior literature is due to the symmetric nature of the surrogate losses used and not due to softmax. We then propose a novel statistically consistent asymmetric softmax-based surrogate loss that can produce valid estimates without the issue of unboundedness. We further analyze the non-asymptotic properties of our proposed method and empirically validate its performance and calibration on benchmark datasets.

PDF Details

NeurIPS Conference 2023 Conference Paper

On the Importance of Feature Separability in Predicting Out-Of-Distribution Error

  • Renchunzi Xie
  • Hongxin Wei
  • Lei Feng
  • Yuzhou Cao
  • Bo An

Estimating the generalization performance is practically challenging on out-of-distribution (OOD) data without ground-truth labels. While previous methods emphasize the connection between distribution difference and OOD accuracy, we show that a large domain gap not necessarily leads to a low test accuracy. In this paper, we investigate this problem from the perspective of feature separability empirically and theoretically. Specifically, we propose a dataset-level score based upon feature dispersion to estimate the test accuracy under distribution shift. Our method is inspired by desirable properties of features in representation learning: high inter-class dispersion and high intra-class compactness. Our analysis shows that inter-class dispersion is strongly correlated with the model accuracy, while intra-class compactness does not reflect the generalization performance on OOD data. Extensive experiments demonstrate the superiority of our method in both prediction performance and computational efficiency.

PDF Details

AAAI Conference 2023 Conference Paper

Partial-Label Regression

  • Xin Cheng
  • Deng-Bao Wang
  • Lei Feng
  • Min-Ling Zhang
  • Bo An

Partial-label learning is a popular weakly supervised learning setting that allows each training example to be annotated with a set of candidate labels. Previous studies on partial-label learning only focused on the classification setting where candidate labels are all discrete, which cannot handle continuous labels with real values. In this paper, we provide the first attempt to investigate partial-label regression, where each training example is annotated with a set of real-valued candidate labels. To solve this problem, we first propose a simple baseline method that takes the average loss incurred by candidate labels as the predictive loss. The drawback of this method lies in that the loss incurred by the true label may be overwhelmed by other false labels. To overcome this drawback, we propose an identification method that takes the least loss incurred by candidate labels as the predictive loss. We further improve it by proposing a progressive identification method to differentiate candidate labels using progressively updated weights for incurred losses. We prove that the latter two methods are model-consistent and provide convergence analysis showing the optimal parametric convergence rate. Our proposed methods are theoretically grounded and can be compatible with any models, optimizers, and losses. Experiments validate the effectiveness of our proposed methods.

PDF Details DOI

IJCAI Conference 2023 Conference Paper

ProMix: Combating Label Noise via Maximizing Clean Sample Utility

  • Ruixuan Xiao
  • Yiwen Dong
  • Haobo Wang
  • Lei Feng
  • Runze Wu
  • Gang Chen
  • Junbo Zhao

Learning with Noisy Labels (LNL) has become an appealing topic, as imperfectly annotated data are relatively cheaper to obtain. Recent state-of-the-art approaches employ specific selection mechanisms to separate clean and noisy samples and then apply Semi-Supervised Learning (SSL) techniques for improved performance. However, the selection step mostly provides a medium-sized and decent-enough clean subset, which overlooks a rich set of clean samples. To fulfill this, we propose a novel LNL framework ProMix that attempts to maximize the utility of clean samples for boosted performance. Key to our method, we propose a matched high confidence selection technique that selects those examples with high confidence scores and matched predictions with given labels to dynamically expand a base clean sample set. To overcome the potential side effect of excessive clean set selection procedure, we further devise a novel SSL framework that is able to train balanced and unbiased classifiers on the separated clean and noisy samples. Extensive experiments demonstrate that ProMix significantly advances the current state-of-the-art results on multiple benchmarks with different types and levels of noise. It achieves an average improvement of 2. 48% on the CIFAR-N dataset.

PDF Details DOI

NeurIPS Conference 2023 Conference Paper

Regression with Cost-based Rejection

  • Xin Cheng
  • Yuzhou Cao
  • Haobo Wang
  • Hongxin Wei
  • Bo An
  • Lei Feng

Learning with rejection is an important framework that can refrain from making predictions to avoid critical mispredictions by balancing between prediction and rejection. Previous studies on cost-based rejection only focused on the classification setting, which cannot handle the continuous and infinite target space in the regression setting. In this paper, we investigate a novel regression problem called regression with cost-based rejection, where the model can reject to make predictions on some examples given certain rejection costs. To solve this problem, we first formulate the expected risk for this problem and then derive the Bayes optimal solution, which shows that the optimal model should reject to make predictions on the examples whose variance is larger than the rejection cost when the mean squared error is used as the evaluation metric. Furthermore, we propose to train the model by a surrogate loss function that considers rejection as binary classification and we provide conditions for the model consistency, which implies that the Bayes optimal solution can be recovered by our proposed surrogate loss. Extensive experiments demonstrate the effectiveness of our proposed method.

PDF Details

NeurIPS Conference 2023 Conference Paper

SPA: A Graph Spectral Alignment Perspective for Domain Adaptation

  • Zhiqing Xiao
  • Haobo Wang
  • Ying Jin
  • Lei Feng
  • Gang Chen
  • Fei Huang
  • Junbo Zhao

Unsupervised domain adaptation (UDA) is a pivotal form in machine learning to extend the in-domain model to the distinctive target domains where the data distributions differ. Most prior works focus on capturing the inter-domain transferability but largely overlook rich intra-domain structures, which empirically results in even worse discriminability. In this work, we introduce a novel graph SPectral Alignment (SPA) framework to tackle the tradeoff. The core of our method is briefly condensed as follows: (i)-by casting the DA problem to graph primitives, SPA composes a coarse graph alignment mechanism with a novel spectral regularizer towards aligning the domain graphs in eigenspaces; (ii)-we further develop a fine-grained message propagation module --- upon a novel neighbor-aware self-training mechanism --- in order for enhanced discriminability in the target domain. On standardized benchmarks, the extensive experiments of SPA demonstrate that its performance has surpassed the existing cutting-edge DA methods. Coupled with dense model analysis, we conclude that our approach indeed possesses superior efficacy, robustness, discriminability, and transferability. Code and data are available at: https: //github. com/CrownX/SPA.

PDF Details

NeurIPS Conference 2022 Conference Paper

Can Adversarial Training Be Manipulated By Non-Robust Features?

  • Lue Tao
  • Lei Feng
  • Hongxin Wei
  • Jinfeng Yi
  • Sheng-Jun Huang
  • Songcan Chen

Adversarial training, originally designed to resist test-time adversarial examples, has shown to be promising in mitigating training-time availability attacks. This defense ability, however, is challenged in this paper. We identify a novel threat model named stability attack, which aims to hinder robust availability by slightly manipulating the training data. Under this threat, we show that adversarial training using a conventional defense budget $\epsilon$ provably fails to provide test robustness in a simple statistical setting, where the non-robust features of the training data can be reinforced by $\epsilon$-bounded perturbation. Further, we analyze the necessity of enlarging the defense budget to counter stability attacks. Finally, comprehensive experiments demonstrate that stability attacks are harmful on benchmark datasets, and thus the adaptive defense is necessary to maintain robustness.

PDF Details

AAAI Conference 2022 Conference Paper

GearNet: Stepwise Dual Learning for Weakly Supervised Domain Adaptation

  • Renchunzi Xie
  • Hongxin Wei
  • Lei Feng
  • Bo An

This paper studies a weakly supervised domain adaptation (WSDA) problem, where we only have access to the source domain with noisy labels, from which we need to transfer useful information to the unlabeled target domain. Although there have been a few studies on this problem, most of them only exploit unidirectional relationships from the source domain to the target domain. In this paper, we propose a universal paradigm called GearNet to exploit bilateral relationships between the two domains. Specifically, we take the two domains as different inputs to train two models alternately, and a symmetrical Kullback-Leibler loss is used for selectively matching the predictions of the two models in the same domain. This interactive learning schema enables implicit label noise canceling and exploit correlations between the source and target domains. Therefore, our GearNet has the great potential to boost the performance of a wide range of existing WSDA methods. Comprehensive experimental results show that the performance of existing methods can be significantly improved by equipping with our GearNet.

PDF Details

NeurIPS Conference 2022 Conference Paper

Generalizing Consistent Multi-Class Classification with Rejection to be Compatible with Arbitrary Losses

  • Yuzhou Cao
  • Tianchi Cai
  • Lei Feng
  • Lihong Gu
  • Jinjie Gu
  • Bo An
  • Gang Niu
  • Masashi Sugiyama

\emph{Classification with rejection} (CwR) refrains from making a prediction to avoid critical misclassification when encountering test samples that are difficult to classify. Though previous methods for CwR have been provided with theoretical guarantees, they are only compatible with certain loss functions, making them not flexible enough when the loss needs to be changed with the dataset in practice. In this paper, we derive a novel formulation for CwR that can be equipped with arbitrary loss functions while maintaining the theoretical guarantees. First, we show that $K$-class CwR is equivalent to a $(K\! +\! 1)$-class classification problem on the original data distribution with an augmented class, and propose an empirical risk minimization formulation to solve this problem with an estimation error bound. Then, we find necessary and sufficient conditions for the learning \emph{consistency} of the surrogates constructed on our proposed formulation equipped with any classification-calibrated multi-class losses, where consistency means the surrogate risk minimization implies the target risk minimization for CwR. Finally, experiments on benchmark datasets validate the effectiveness of our proposed method.

PDF Details

TMLR Journal 2022 Journal Article

SemiNLL: A Framework of Noisy-Label Learning by Semi-Supervised Learning

  • Zhuowei Wang
  • Jing Jiang
  • Bo Han
  • Lei Feng
  • Bo An
  • Gang Niu
  • Guodong Long

Deep learning with noisy labels is a challenging task, which has received much attention from the machine learning and computer vision communities. Recent prominent methods that build on a specific sample selection (SS) strategy and a specific semi-supervised learning (SSL) model achieved state-of-the-art performance. Intuitively, better performance could be achieved if stronger SS strategies and SSL models are employed. Following this intuition, one might easily derive various effective noisy-label learning methods using different combinations of SS strategies and SSL models, which is, however, simply reinventing the wheel in essence. To prevent this problem, we propose SemiNLL, a versatile framework that investigates how to naturally combine different SS and SSL components based on their effects and efficiencies. We conduct a systematic and detailed analysis of the combinations of possible components based on our framework. Our framework can absorb various SS strategies and SSL backbones, utilizing their power to achieve promising performance. The instantiations of our framework demonstrate substantial improvements over state-of-the-art methods on benchmark-simulated and real-world datasets with noisy labels.

PDF Details

NeurIPS Conference 2022 Conference Paper

SoLar: Sinkhorn Label Refinery for Imbalanced Partial-Label Learning

  • Haobo Wang
  • Mingxuan Xia
  • Yixuan Li
  • Yuren Mao
  • Lei Feng
  • Gang Chen
  • Junbo Zhao

Partial-label learning (PLL) is a peculiar weakly-supervised learning task where the training samples are generally associated with a set of candidate labels instead of single ground truth. While a variety of label disambiguation methods have been proposed in this domain, they normally assume a class-balanced scenario that may not hold in many real-world applications. Empirically, we observe degenerated performance of the prior methods when facing the combinatorial challenge from the long-tailed distribution and partial-labeling. In this work, we first identify the major reasons that the prior work failed. We subsequently propose SoLar, a novel Optimal Transport-based framework that allows to refine the disambiguated labels towards matching the marginal class prior distribution. SoLar additionally incorporates a new and systematic mechanism for estimating the long-tailed class prior distribution under the PLL setup. Through extensive experiments, SoLar exhibits substantially superior results on standardized benchmarks compared to the previous state-of-the-art PLL methods. Code and data are available at: https: //github. com/hbzju/SoLar.

PDF Details

AAAI Conference 2022 Conference Paper

With False Friends Like These, Who Can Notice Mistakes?

  • Lue Tao
  • Lei Feng
  • Jinfeng Yi
  • Songcan Chen

Adversarial examples crafted by an explicit adversary have attracted significant attention in machine learning. However, the security risk posed by a potential false friend has been largely overlooked. In this paper, we unveil the threat of hypocritical examples—inputs that are originally misclassified yet perturbed by a false friend to force correct predictions. While such perturbed examples seem harmless, we point out for the first time that they could be maliciously used to conceal the mistakes of a substandard (i. e. , not as good as required) model during an evaluation. Once a deployer trusts the hypocritical performance and applies the “well-performed” model in realworld applications, unexpected failures may happen even in benign environments. More seriously, this security risk seems to be pervasive: we find that many types of substandard models are vulnerable to hypocritical examples across multiple datasets. Furthermore, we provide the first attempt to characterize the threat with a metric called hypocritical risk and try to circumvent it via several countermeasures. Results demonstrate the effectiveness of the countermeasures, while the risk remains non-negligible even after adaptive robust training.

PDF Details

NeurIPS Conference 2021 Conference Paper

Better Safe Than Sorry: Preventing Delusive Adversaries with Adversarial Training

  • Lue Tao
  • Lei Feng
  • Jinfeng Yi
  • Sheng-Jun Huang
  • Songcan Chen

Delusive attacks aim to substantially deteriorate the test accuracy of the learning model by slightly perturbing the features of correctly labeled training examples. By formalizing this malicious attack as finding the worst-case training data within a specific $\infty$-Wasserstein ball, we show that minimizing adversarial risk on the perturbed data is equivalent to optimizing an upper bound of natural risk on the original data. This implies that adversarial training can serve as a principled defense against delusive attacks. Thus, the test accuracy decreased by delusive attacks can be largely recovered by adversarial training. To further understand the internal mechanism of the defense, we disclose that adversarial training can resist the delusive perturbations by preventing the learner from overly relying on non-robust features in a natural setting. Finally, we complement our theoretical findings with a set of experiments on popular benchmark datasets, which show that the defense withstands six different practical attacks. Both theoretical and empirical results vote for adversarial training when confronted with delusive adversaries.

PDF Details

IS Journal 2021 Journal Article

Embedding-Augmented Generalized Matrix Factorization for Recommendation With Implicit Feedback

  • Lei Feng
  • Hongxin Wei
  • Qingyu Guo
  • Zhuoyi Lin
  • Bo An

Learning effective representations of users and items is crucially important to recommendation with implicit feedback. Matrix factorization is the basic idea to derive the representations of users and items by decomposing the given interaction matrix. However, existing matrix factorization based approaches share the limitation in that the interaction between user embedding and item embedding is only weakly enforced by fitting the given individual rating value, which may lose potentially useful information. In this article, we propose a novel augmented generalized matrix factorization approach that is able to incorporate the historical interaction information of users and items for learning effective representations of users and items. Despite the simplicity of our proposed approach, extensive experiments on four public implicit feedback datasets demonstrate that our approach outperforms state-of-the-art counterparts. Furthermore, the ablation study demonstrates that by using the historical interactions to enrich user embedding and item embedding for generalized matrix factorization, better performance, faster convergence, and lower training loss can be achieved.

Details DOI

IJCAI Conference 2021 Conference Paper

Learning from Complementary Labels via Partial-Output Consistency Regularization

  • Deng-Bao Wang
  • Lei Feng
  • Min-Ling Zhang

In complementary-label learning (CLL), a multi-class classifier is learned from training instances each associated with complementary labels, which specify the classes that the instance does not belong to. Previous studies focus on unbiased risk estimator or surrogate loss while neglect the importance of regularization in training phase. In this paper, we give the first attempt to leverage regularization techniques for CLL. By decoupling a label vector into complementary labels and partial unknown labels, we simultaneously inhibit the outputs of complementary labels with a complementary loss and penalize the sensitivity of the classifier on the partial outputs of these unknown classes by consistency regularization. Then we unify the complementary loss and consistency loss together by a specially designed dynamic weighting factor. We conduct a series of experiments showing that the proposed method achieves highly competitive performance in CLL.

PDF Details DOI

NeurIPS Conference 2021 Conference Paper

Rethinking Calibration of Deep Neural Networks: Do Not Be Afraid of Overconfidence

  • Deng-Bao Wang
  • Lei Feng
  • Min-Ling Zhang

Capturing accurate uncertainty quantification of the prediction from deep neural networks is important in many real-world decision-making applications. A reliable predictor is expected to be accurate when it is confident about its predictions and indicate high uncertainty when it is likely to be inaccurate. However, modern neural networks have been found to be poorly calibrated, primarily in the direction of overconfidence. In recent years, there is a surge of research on model calibration by leveraging implicit or explicit regularization techniques during training, which obtain well calibration by avoiding overconfident outputs. In our study, we empirically found that despite the predictions obtained from these regularized models are better calibrated, they suffer from not being as calibratable, namely, it is harder to further calibrate their predictions with post-hoc calibration methods like temperature scaling and histogram binning. We conduct a series of empirical studies showing that overconfidence may not hurt final calibration performance if post-hoc calibration is allowed, rather, the penalty of confident outputs will compress the room of potential improvements in post-hoc calibration phase. Our experimental findings point out a new direction to improve calibration of DNNs by considering main training and post-hoc calibration as a unified framework.

PDF Details

IJCAI Conference 2020 Conference Paper

Can Cross Entropy Loss Be Robust to Label Noise?

  • Lei Feng
  • Senlin Shu
  • Zhuoyi Lin
  • Fengmao Lv
  • Li Li
  • Bo An

Trained with the standard cross entropy loss, deep neural networks can achieve great performance on correctly labeled data. However, if the training data is corrupted with label noise, deep models tend to overfit the noisy labels, thereby achieving poor generation performance. To remedy this issue, several loss functions have been proposed and demonstrated to be robust to label noise. Although most of the robust loss functions stem from Categorical Cross Entropy (CCE) loss, they fail to embody the intrinsic relationships between CCE and other loss functions. In this paper, we propose a general framework dubbed Taylor cross entropy loss to train deep models in the presence of label noise. Specifically, our framework enables to weight the extent of fitting the training labels by controlling the order of Taylor Series for CCE, hence it can be robust to label noise. In addition, our framework clearly reveals the intrinsic relationships between CCE and other loss functions, such as Mean Absolute Error (MAE) and Mean Squared Error (MSE). Moreover, we present a detailed theoretical analysis to certify the robustness of this framework. Extensive experimental results on benchmark datasets demonstrate that our proposed approach significantly outperforms the state-of-the-art counterparts.

PDF Details DOI

IJCAI Conference 2020 Conference Paper

Discovering Latent Class Labels for Multi-Label Learning

  • Jun Huang
  • Linchuan Xu
  • Jing Wang
  • Lei Feng
  • Kenji Yamanishi

Existing multi-label learning (MLL) approaches mainly assume all the labels are observed and construct classification models with a fixed set of target labels (known labels). However, in some real applications, multiple latent labels may exist outside this set and hide in the data, especially for large-scale data sets. Discovering and exploring the latent labels hidden in the data may not only find interesting knowledge but also help us to build a more robust learning model. In this paper, a novel approach named DLCL (i. e. , Discovering Latent Class Labels for MLL) is proposed which can not only discover the latent labels in the training data but also predict new instances with the latent and known labels simultaneously. Extensive experiments show a competitive performance of DLCL against other state-of-the-art MLL approaches.

PDF Details DOI

NeurIPS Conference 2020 Conference Paper

Provably Consistent Partial-Label Learning

  • Lei Feng
  • Jiaqi Lv
  • Bo Han
  • Miao Xu
  • Gang Niu
  • Xin Geng
  • Bo An
  • Masashi Sugiyama

Partial-label learning (PLL) is a multi-class classification problem, where each training example is associated with a set of candidate labels. Even though many practical PLL methods have been proposed in the last two decades, there lacks a theoretical understanding of the consistency of those methods - none of the PLL methods hitherto possesses a generation process of candidate label sets, and then it is still unclear why such a method works on a specific dataset and when it may fail given a different dataset. In this paper, we propose the first generation model of candidate label sets, and develop two PLL methods that are guaranteed to be provably consistent, i. e. , one is risk-consistent and the other is classifier-consistent. Our methods are advantageous, since they are compatible with any deep network or stochastic optimizer. Furthermore, thanks to the generation model, we would be able to answer the two questions above by testing if the generation model matches given candidate label sets. Experiments on benchmark and real-world datasets validate the effectiveness of the proposed generation model and two PLL methods.

PDF Details

AAAI Conference 2019 Conference Paper

Collaboration Based Multi-Label Learning

  • Lei Feng
  • Bo An
  • Shuo He

It is well-known that exploiting label correlations is crucially important to multi-label learning. Most of the existing approaches take label correlations as prior knowledge, which may not correctly characterize the real relationships among labels. Besides, label correlations are normally used to regularize the hypothesis space, while the final predictions are not explicitly correlated. In this paper, we suggest that for each individual label, the final prediction involves the collaboration between its own prediction and the predictions of other labels. Based on this assumption, we first propose a novel method to learn the label correlations via sparse reconstruction in the label space. Then, by seamlessly integrating the learned label correlations into model training, we propose a novel multi-label learning approach that aims to explicitly account for the correlated predictions of labels while training the desired model simultaneously. Extensive experimental results show that our approach outperforms the state-of-the-art counterparts.

PDF Details

IJCAI Conference 2019 Conference Paper

Partial Label Learning by Semantic Difference Maximization

  • Lei Feng
  • Bo An

Partial label learning is a weakly supervised learning framework, in which each instance is provided with multiple candidate labels while only one of them is correct. Most of the existing approaches focus on leveraging the instance relationships to disambiguate the given noisy label space, while it is still unclear whether we can exploit potentially useful information in label space to alleviate the label ambiguities. This paper gives a positive answer to this question for the first time. Specifically, if two instances do not share any common candidate labels, they cannot have the same ground-truth label. By exploiting such dissimilarity relationships from label space, we propose a novel approach that aims to maximize the latent semantic differences of the two instances whose ground-truth labels are definitely different, while training the desired model simultaneously, thereby continually enlarging the gap of label confidences between two instances of different classes. Extensive experiments on artificial and real-world partial label datasets show that our approach significantly outperforms state-of-the-art counterparts.

PDF Details

AAAI Conference 2019 Conference Paper

Partial Label Learning with Self-Guided Retraining

  • Lei Feng
  • Bo An

Partial label learning deals with the problem where each training instance is assigned a set of candidate labels, only one of which is correct. This paper provides the first attempt to leverage the idea of self-training for dealing with partially labeled examples. Specifically, we propose a unified formulation with proper constraints to train the desired model and perform pseudo-labeling jointly. For pseudo-labeling, unlike traditional self-training that manually differentiates the ground-truth label with enough high confidence, we introduce the maximum infinity norm regularization on the modeling outputs to automatically achieve this consideratum, which results in a convex-concave optimization problem. We show that optimizing this convex-concave problem is equivalent to solving a set of quadratic programming (QP) problems. By proposing an upper-bound surrogate objective function, we turn to solving only one QP problem for improving the optimization efficiency. Extensive experiments on synthesized and real-world datasets demonstrate that the proposed approach significantly outperforms the state-of-the-art partial label learning approaches.

PDF Details

IJCAI Conference 2018 Conference Paper

Leveraging Latent Label Distributions for Partial Label Learning

  • Lei Feng
  • Bo An

In partial label learning, each training example is assigned a set of candidate labels, only one of which is the ground-truth label. Existing partial label learning frameworks either assume each candidate label of equal confidence or consider the ground-truth label as a latent variable hidden in the indiscriminate candidate label set, while the different labeling confidence levels of the candidate labels are regrettably ignored. In this paper, we formalize the different labeling confidence levels as the latent label distributions, and propose a novel unified framework to estimate the latent label distributions while training the model simultaneously. Specifically, we present a biconvex formulation with constrained local consistency and adopt an alternating method to solve this optimization problem. The process of alternating optimization exactly facilitates the mutual adaption of the model training and the constrained label propagation. Extensive experimental results on controlled UCI datasets as well as real-world datasets clearly show the effectiveness of the proposed approach.

PDF Details