Author name cluster

Nan Pu

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.

5 papers

2 author rows

AAAI Conference 2026 Conference Paper

Open-World Deepfake Attribution via Confidence-Aware Asymmetric Learning

Haiyang Zheng
Nan Pu
Wenjing Li
Teng Long
Nicu Sebe
Zhun Zhong

The proliferation of synthetic facial imagery has intensified the need for robust Open-World DeepFake Attribution (OW-DFA), which aims to attribute both known and unknown forgeries using labeled data for known types and unlabeled data containing a mixture of known and novel types. However, existing OW-DFA methods face two critical limitations: 1) A confidence skew that leads to unreliable pseudo-labels for novel forgeries, resulting in biased training. 2) An unrealistic assumption that the number of unknown forgery types is known a priori. To address these challenges, we propose a Confidence-aware Asymmetric Learning (CAL) framework, which adaptively balances model confidence across known and novel forgery types. CAL mainly consists of two components: Confidence-aware Consistency Regularization (CCR) and Asymmetric Confidence Reinforcement (ACR). CCR mitigates pseudo-label bias by dynamically scaling sample losses based on normalized confidence, gradually shifting the training focus from high- to low-confidence samples. ACR complements this by separately calibrating confidence for known and novel classes through selective learning on high-confidence samples, guided by their confidence gap. Together, CCR and ACR form a mutually reinforcing loop that significantly improves the model's OW-DFA performance. Moreover, we introduce a Dynamic Prototype Pruning (DPP) strategy that automatically estimates the number of novel forgery types in a coarse-to-fine manner, removing the need for unrealistic prior assumptions and enhancing the scalability of our methods to real-world OW-DFA scenarios. Extensive experiments on the standard and OW-DFA benchmark and a newly extended benchmark incorporating advanced manipulations demonstrate that CAL consistently outperforms previous methods, achieving new state-of-the-art performance on both known and novel forgery attribution.

PDF Details DOI

TMLR Journal 2025 Journal Article

Federated Generalized Novel Category Discovery with Prompts Tuning

Lei Shen
Nan Pu
Zhun Zhong
Mingming Gong
Dianhai Yu
Chengqi Zhang
Bo Han

Generalized category discovery (GCD) is proposed to handle categories from unseen labels during the inference stage by clustering them. Most works in GCD provide solutions for unseen classes in data-centralized settings. However, unlabeled categories possessed by clients, which are common in real-world federated learning (FL), have been largely ignored and degraded the performance of classic FL algorithms. To demonstrate and mitigate the harmful effect of unseen classes, we dive into a GCD problem setting applicable for FL named FedGCD, analyze overfitting problem in FedGCD in detail, establish a strong baseline constructed with state-of-the-art GCD algorithm simGCD, and design a learning framework with prompt tuning to tackle both the overfitting and communication burden problems in FedGCD. In our methods, clients first separately carry out prompt learning on local data. Then, we aggregate the prompts from all clients as the global prompt to help capture global knowledge and then send the global prompts to local clients to allow access to broader knowledge from other clients. By this method, we significantly reduce the parameters needed to upload in FedGCD, which is a common obstacle in the real application of most FL algorithms. We conduct experiments on both generic and fine-grained datasets like CIFAR-100 and CUB-200, and show that our method is comparable to the FL version of simGCD and surpasses other baselines with significantly fewer parameters to transmit.

PDF Details

ICML Conference 2025 Conference Paper

Robust Consensus Anchor Learning for Efficient Multi-view Subspace Clustering

Yalan Qin
Nan Pu
Guorui Feng
Nicu Sebe

As a leading unsupervised classification algorithm in artificial intelligence, multi-view subspace clustering segments unlabeled data from different subspaces. Recent works based on the anchor have been proposed to decrease the computation complexity for the datasets with large scales in multi-view clustering. The major differences among these methods lie on the objective functions they define. Despite considerable success, these works pay few attention to guaranting the robustness of learned consensus anchors via effective manner for efficient multi-view clustering and investigating the specific local distribution of cluster in the affine subspace. Besides, the robust consensus anchors as well as the common cluster structure shared by different views are not able to be simultaneously learned. In this paper, we propose Robust Consensus anchors learning for efficient multi-view Subspace Clustering (RCSC). We first show that if the data are sufficiently sampled from independent subspaces, and the objective function meets some conditions, the achieved anchor graph has the block-diagonal structure. As a special case, we provide a model based on Frobenius norm, non-negative and affine constraints in consensus anchors learning, which guarantees the robustness of learned consensus anchors for efficient multi-view clustering and investigates the specific local distribution of cluster in the affine subspace. While it is simple, we theoretically give the geometric analysis regarding the formulated RCSC. The union of these three constraints is able to restrict how each data point is described in the affine subspace with specific local distribution of cluster for guaranting the robustness of learned consensus anchors. RCSC takes full advantages of correlation among consensus anchors, which encourages the grouping effect and groups highly correlated consensus anchors together with the guidance of view-specific projection. The anchor graph construction, partition and robust anchor learning are jointly integrated into a unified framework. It ensures the mutual enhancement for these procedures and helps lead to more discriminative consensus anchors as well as the cluster indicator. We then adopt an alternative optimization strategy for solving the formulated problem. Experiments performed on eight multi-view datasets confirm the superiority of RCSC based on the effectiveness and efficiency.

Details

NeurIPS Conference 2024 Conference Paper

Prototypical Hash Encoding for On-the-Fly Fine-Grained Category Discovery

Haiyang Zheng
Nan Pu
Wenjing Li
Nicu Sebe
Zhun Zhong

In this paper, we study a practical yet challenging task, On-the-fly Category Discovery (OCD), aiming to online discover the newly-coming stream data that belong to both known and unknown classes, by leveraging only known category knowledge contained in labeled data. Previous OCD methods employ the hash-based technique to represent old/new categories by hash codes for instance-wise inference. However, directly mapping features into low-dimensional hash space not only inevitably damages the ability to distinguish classes and but also causes ``high sensitivity'' issue, especially for fine-grained classes, leading to inferior performance. To address these drawbacks, we propose a novel Prototypical Hash Encoding (PHE) framework consisting of Category-aware Prototype Generation (CPG) and Discriminative Category Encoding (DCE) to mitigate the sensitivity of hash code while preserving rich discriminative information contained in high-dimension feature space, in a two-stage projection fashion. CPG enables the model to fully capture the intra-category diversity by representing each category with multiple prototypes. DCE boosts the discrimination ability of hash code with the guidance of the generated category prototypes and the constraint of minimum separation distance. By jointly optimizing CPG and DCE, we demonstrate that these two components are mutually beneficial towards an effective OCD. Extensive experiments show the significant superiority of our PHE over previous methods, e. g. obtaining an improvement of +5. 3% in ALL ACC averaged on all datasets. Moreover, due to the nature of the interpretable prototypes, we visually analyze the underlying mechanism of how PHE helps group certain samples into either known or unknown categories. Code is available at https: //github. com/HaiyangZheng/PHE.

PDF Details DOI

AAAI Conference 2023 Conference Paper

COCA: COllaborative CAusal Regularization for Audio-Visual Question Answering

Mingrui Lao
Nan Pu
Yu Liu
Kai He
Erwin M. Bakker
Michael S. Lew

Audio-Visual Question Answering (AVQA) is a sophisticated QA task, which aims at answering textual questions over given video-audio pairs with comprehensive multimodal reasoning. Through detailed causal-graph analyses and careful inspections of their learning processes, we reveal that AVQA models are not only prone to over-exploit prevalent language bias, but also suffer from additional joint-modal biases caused by the shortcut relations between textual-auditory/visual co-occurrences and dominated answers. In this paper, we propose a COllabrative CAusal (COCA) Regularization to remedy this more challenging issue of data biases. Specifically, a novel Bias-centered Causal Regularization (BCR) is proposed to alleviate specific shortcut biases by intervening bias-irrelevant causal effects, and further introspect the predictions of AVQA models in counterfactual and factual scenarios. Based on the fact that the dominated bias impairing model robustness for different samples tends to be different, we introduce a Multi-shortcut Collaborative Debiasing (MCD) to measure how each sample suffers from different biases, and dynamically adjust their debiasing concentration to different shortcut correlations. Extensive experiments demonstrate the effectiveness as well as backbone-agnostic ability of our COCA strategy, and it achieves state-of-the-art performance on the large-scale MUSIC-AVQA dataset.

PDF Details DOI