Author name cluster

Shidong Wang

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.

6 papers

1 author row

AAAI Conference 2026 Conference Paper

Foundation-Adaptive Integrated Refinement for Generalized Category Discovery

Yuwei Bian
Shidong Wang
Yazhou Yao
Haofeng Zhang

The potential of Generalized Category Discovery (GCD) lies in its ability to identify previously undiscovered patterns in both labeled and unlabeled data by leveraging insights from partially labeled training samples. However, interference can arise due to the model's dual focus on discovering both novel and known categories, often leading to conflicts that obscure true patterns in the dataset. This paper presents a divide-and-conquer framework, Foundation-Adaptive Integrated Refinement (FAIR), which fine-tunes pretrained foundational weights for various purposes, divided into Foundation (pretrained weights), Adaptive (weights fine-tuned with a variance-preserving loss), and Integrated (weights adjusted for both labeled and unlabeled data). The Adaptive utilizes a newly proposed adaptive contrastive loss that introduces variances within classes to preserve the individuality of representations. The Integrated addresses inherent estimation errors while dynamically estimating the number of categories, incorporating a cosine-based perturbation mechanism as a relaxed margin to accommodate potential ground-truth deviations, rather than relying on biased estimates. Extensive experiments on six benchmark datasets demonstrate our method's effectiveness, outperforming state-of-the-art algorithms, especially on fine-grained datasets.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Learning a Fix and Explore Framework for Continuous Generalized Category Discovery

Chunming Li
Shidong Wang
Haofeng Zhang

To address the limitations of transductive learning in evolving real-world scenarios where unknown categories may continuously emerge, Continual Generalized Category Discovery (C-GCD) presents a novel paradigm that extends conventional category discovery frameworks. Unlike traditional static learning environments, C-GCD requires models to incrementally discover novel categories across multiple operational phases while maintaining discrimination capabilities for previously learned classes, posing significant challenges in balancing stability and plasticity. Prior approaches typically employ parameter-level knowledge distillation from historical models to alleviate catastrophic forgetting, which effectively preserves prior knowledge and optimizes computational efficiency. However, our analysis reveals that the persistent availability of samples from previous stages enables more sophisticated knowledge preservation strategies. Specifically, we present a Fix and Explore strategy that employs distinct learning methodologies for different types of potential data, aiming to preserve the features of old categories as much as possible and gradually exploring the potential distribution of new class latent spaces, we can enhance the model's ability to discover novel categories. This paper investigates this effect and introduces a novel heuristic paradigm to solve the C-GCD problem, called Fix and Explore (FaE), which aims to provide sufficient imaginative space for new classes while preserving the classification ability for old tasks. We conducted experiments across multiple datasets and performed detailed comparisons. The results demonstrate that our method achieves state-of-the-art performance at each stage across all datasets.

PDF Details DOI

IJCAI Conference 2025 Conference Paper

Few-shot Novel Category Discovery

Chunming Li
Shidong Wang
Haofeng Zhang

The recently proposed Novel Category Discovery (NCD) adapt paradigm of transductive learning hinders its application in more real-world scenarios. In fact, few labeled data in part of new categories can well alleviate this burden, which coincides with the ease that people can label few of new category data. Therefore, this paper presents a new setting in which a trained agent is able to flexibly switch between the tasks of identifying examples of known (labelled) classes and clustering novel (completely unlabeled) classes as the number of query examples increases by leveraging knowledge learned from only a few (handful) support examples. Drawing inspiration from the discovery of novel categories using prior-based clustering algorithms, we introduce a novel framework that further relaxes its assumptions to the real-world open set level by unifying the concept of model adaptability in few-shot learning. We refer to this setting as Few-Shot Novel Category Discovery (FSNCD) and propose Semi-supervised Hierarchical Clustering (SHC) and Uncertainty-aware K-means Clustering (UKC) to examine the model's reasoning capabilities. Extensive experiments and detailed analysis on five commonly used datasets demonstrate that our methods can achieve leading performance levels across different task settings and scenarios. Code is available at: https: //github. com/Ashengl/FSNCD.

PDF Details DOI

AIIM Journal 2025 Journal Article

RobustEMD: Domain robust matching for cross-domain few-shot medical image segmentation

Yazhou Zhu
Minxian Li
Qiaolin Ye
Shidong Wang
Tong Xin
Haofeng Zhang

Few-shot medical image segmentation (FSMIS) aims to perform the limited annotated data learning in the medical image analysis scope. Despite the progress has been achieved, current FSMIS models are all trained and deployed on the same data domain, as is not consistent with the clinical reality that medical imaging data is always across different data domains (e. g. imaging modalities, institutions and equipment sequences). In this paper, we introduce Cross-domain Few-shot Medical Image Segmentation (CD-FSMIS) and propose a RobustEMD matching mechanism based on Earth Mover’s Distance (EMD) to enhance cross-domain generalization. Our approach includes three key components: (1) a channel-wise feature decomposition strategy that uniformly divides support and query features into local nodes, (2) a texture structure aware weights generation method that restrains domain-relevant nodes through Sobel-based gradient calculation, and (3) a boundary-aware Hausdorff distance measurement for transportation cost calculation. Extensive experiments across three scenarios (cross-modal, cross-sequence and cross-institution) show that our method significantly outperforms existing approaches. And ablation studies further confirm that each component of our RobustEMD mechanism contributes to the enhanced performance. The experimental outcomes highlight strong generalization capabilities of our model in real-world heterogeneous medical imaging environments. Code is available at https: //github. com/YazhouZhu19/RobustEMD.

Details DOI

TIST Journal 2024 Journal Article

Cascaded Alternating Refinement Transformer for Few-shot Medical Image Segmentation

Ziming Cheng
Yazhou Zhu
Shidong Wang
Tong Xin
Haofeng Zhang

Conventional biomedical image segmentation heavily relies on substantial annotations, which demand significant human and financial resources for collection. Consequently, learning a model with excellent performance using limited medical image data becomes a challenging problem. Upliftingly, the advent of few-shot medical image segmentation (FSMIS) offers a potential solution. Although prototypical networks are commonly employed in existing FSMIS tasks, the prototypes derived from support features often induce significant bias issues caused by intra-class variations. To this end, we propose a method called Cascaded Altering Refinement Transformer (CART) to iteratively calibrate the prototypes with both support and query features. This method focuses on capturing the commonality between foreground information of the support and query features using the Alterating Refinement Transformer (ART) module, which includes two Multi-head Cross Attention (MCA) modules. Furthermore, we cascade ART modules to refine the class prototypes, resulting in representative prototypes. This process ultimately contributes to a more accurate predicted mask. Besides, to preserve more valid information in each cascaded ART module and achieve better performance, we propose a novel inference method that accumulates the predicted segmentation masks in all ART modules by applying the Rounding-Up strategy. Extensive experiments on three public medical image datasets demonstrate that our model outperforms the state-of-the-art methods, and detailed analysis also validates the reasonableness of this design. Code is available at: https://github.com/zmcheng9/CART.

Details DOI

AAAI Conference 2022 Conference Paper

Boosting Generative Zero-Shot Learning by Synthesizing Diverse Features with Attribute Augmentation

Xiaojie Zhao
Yuming Shen
Shidong Wang
Haofeng Zhang

The recent advance in deep generative models outlines a promising perspective in the realm of Zero-Shot Learning (ZSL). Most generative ZSL methods use category semantic attributes plus a Gaussian noise to generate visual features. After generating unseen samples, this family of approaches effectively transforms the ZSL problem into a supervised classification scheme. However, the existing models use a single semantic attribute, which contains the complete attribute information of the category. The generated data also carry the complete attribute information, but in reality, visual samples usually have limited attributes. Therefore, the generated data from attribute could have incomplete semantics. Based on this fact, we propose a novel framework to boost ZSL by synthesizing diverse features. This method uses augmented semantic attributes to train the generative model, so as to simulate the real distribution of visual features. We evaluate the proposed model on four benchmark datasets, observing significant performance improvement against the state-ofthe-art.

PDF Details