Author name cluster

Yinghuan Shi

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.

18 papers

2 author rows

AAAI Conference 2026 Conference Paper

Decomposing and Composing: Towards Efficient Vision-Language Continual Learning via Rank-1 Expert Pool in a Single LoRA

Zhan Fa
Yue Duan
Jian Zhang
Lei Qi
Wanqi Yang
Yinghuan Shi

Continual learning (CL) in vision-language models (VLMs) faces significant challenges in improving task adaptation and avoiding catastrophic forgetting. Existing methods usually have heavy inference burden or rely on external knowledge, while Low-Rank Adaptation (LoRA) has shown potential in reducing these issues by enabling parameter-efficient tuning. However, considering directly using LoRA to alleviate the catastrophic forgetting problem is non-trivial, we introduce a novel framework that restructures a single LoRA module as a decomposable Rank-1 Expert Pool. Our method learns to dynamically compose a sparse, task-specific update by selecting from this expert pool, guided by the semantics of the [CLS] token. In addition, we propose an Activation-Guided Orthogonal (AGO) loss that orthogonalizes critical parts of LoRA weights across tasks. This sparse composition and orthogonalization enable fewer parameter updates, resulting in domain-aware learning while minimizing inter-task interference and maintaining downstream task performance. Extensive experiments across multiple settings demonstrate state-of-the-art results in all metrics, surpassing zero-shot upper bounds in generalization. Notably, it reduces trainable parameters by 96.7% compared to the baseline method, eliminating reliance on external datasets or task-ID discriminators. The merged LoRAs retain less weights and incur no inference latency, making our method computationally lightweight.

PDF Details DOI

ICLR Conference 2025 Conference Paper

Leveraging Flatness to Improve Information-Theoretic Generalization Bounds for SGD

Ze Peng 0001
Jian Zhang 0090
Yisen Wang 0001
Lei Qi 0001
Yinghuan Shi
Yang Gao 0001

Information-theoretic (IT) generalization bounds have been used to study the generalization of learning algorithms. These bounds are intrinsically data- and algorithm-dependent so that one can exploit the properties of data and algorithm to derive tighter bounds. However, we observe that although the flatness bias is crucial for SGD’s generalization, these bounds fail to capture the improved generalization under better flatness and are also numerically loose. This is caused by the inadequate leverage of SGD's flatness bias in existing IT bounds. This paper derives a more flatness-leveraging IT bound for the flatness-favoring SGD. The bound indicates the learned models generalize better if the large-variance directions of the final weight covariance have small local curvatures in the loss landscape. Experiments on deep neural networks show our bound not only correctly reflects the better generalization when flatness is improved, but is also numerically much tighter. This is achieved by a flexible technique called "omniscient trajectory". When applied to Gradient Descent’s minimax excess risk on convex-Lipschitz-Bounded problems, it improves representative IT bounds’ $\Omega(1)$ rates to $O(1/\sqrt{n})$. It also implies a by-pass of memorization-generalization trade-offs. Codes are available at [https://github.com/peng-ze/omniscient-bounds](https://github.com/peng-ze/omniscient-bounds).

Details

AAAI Conference 2025 Conference Paper

Text and Image Are Mutually Beneficial: Enhancing Training-Free Few-Shot Classification with CLIP

Yayuan Li
Jintao Guo
Lei Qi
Wenbin Li
Yinghuan Shi

Contrastive Language-Image Pretraining (CLIP) has been widely used in vision tasks. Notably, CLIP has demonstrated promising performance in few-shot learning (FSL). However, existing CLIP-based methods in training-free FSL (i.e., without the requirement of additional training) mainly learn different modalities independently, leading to two essential issues: 1) severe anomalous match in image modality; 2) varying quality of generated text prompts. To address these issues, we build a mutual guidance mechanism, that introduces an Image-Guided-Text (IGT) component to rectify varying quality of text prompts through image representations, and a Text-Guided-Image (TGI) component to mitigate the anomalous match of image modality through text representations. By integrating IGT and TGI, we adopt a perspective of Text-Image Mutual guidance Optimization, proposing TIMO. Extensive experiments show that TIMO significantly outperforms the state-of-the-art (SOTA) training-free method. Additionally, by exploring the extent of mutual guidance, we propose an enhanced variant, TIMO-S, which even surpasses the best training-required methods by 0.33% with approximately ×100 less time cost.

PDF Details DOI

JBHI Journal 2024 Journal Article

ERNet: Edge Regularization Network for Cerebral Vessel Segmentation in Digital Subtraction Angiography Images

Weijin Xu
Huihua Yang
Yinghuan Shi
Tao Tan
Wentao Liu
Xipeng Pan
Yiming Deng
Feng Gao

Stroke is a leading cause of disability and fatality in the world, with ischemic stroke being the most common type. Digital Subtraction Angiography images, the gold standard in the operation process, can accurately show the contours and blood flow of cerebral vessels. The segmentation of cerebral vessels in DSA images can effectively help physicians assess the lesions. However, due to the disturbances in imaging parameters and changes in imaging scale, accurate cerebral vessel segmentation in DSA images is still a challenging task. In this paper, we propose a novel Edge Regularization Network (ERNet) to segment cerebral vessels in DSA images. Specifically, ERNet employs the erosion and dilation processes on the original binary vessel annotation to generate pseudo-ground truths of False Negative and False Positive, which serve as constraints to refine the coarse predictions based on their mapping relationship with the original vessels. In addition, we exploit a Hybrid Fusion Module based on convolution and transformers to extract local features and build long-range dependencies. Moreover, to support and advance the open research in the field of ischemic stroke, we introduce FPDSA, the first pixel-level semantic segmentation dataset for cerebral vessels. Extensive experiments on FPDSA illustrate the leading performance of our ERNet.

Details DOI

AAAI Conference 2024 Conference Paper

PG-LBO: Enhancing High-Dimensional Bayesian Optimization with Pseudo-Label and Gaussian Process Guidance

Taicai Chen
Yue Duan
Dong Li
Lei Qi
Yinghuan Shi
Yang Gao

Variational Autoencoder based Bayesian Optimization (VAE-BO) has demonstrated its excellent performance in addressing high-dimensional structured optimization problems. However, current mainstream methods overlook the potential of utilizing a pool of unlabeled data to construct the latent space, while only concentrating on designing sophisticated models to leverage the labeled data. Despite their effective usage of labeled data, these methods often require extra network structures, additional procedure, resulting in computational inefficiency. To address this issue, we propose a novel method to effectively utilize unlabeled data with the guidance of labeled data. Specifically, we tailor the pseudo-labeling technique from semi-supervised learning to explicitly reveal the relative magnitudes of optimization objective values hidden within the unlabeled data. Based on this technique, we assign appropriate training weights to unlabeled data to enhance the construction of a discriminative latent space. Furthermore, we treat the VAE encoder and the Gaussian Process (GP) in Bayesian optimization as a unified deep kernel learning process, allowing the direct utilization of labeled data, which we term as Gaussian Process guidance. This directly and effectively integrates the goal of improving GP accuracy into the VAE training, thereby guiding the construction of the latent space. The extensive experiments demonstrate that our proposed method outperforms existing VAE-BO algorithms in various optimization scenarios. Our code will be published at https://github.com/TaicaiChen/PG-LBO.

PDF Details DOI

AAAI Conference 2024 Conference Paper

Roll with the Punches: Expansion and Shrinkage of Soft Label Selection for Semi-supervised Fine-Grained Learning

Yue Duan
Zhen Zhao
Lei Qi
Luping Zhou
Lei Wang
Yinghuan Shi

While semi-supervised learning (SSL) has yielded promising results, the more realistic SSL scenario remains to be explored, in which the unlabeled data exhibits extremely high recognition difficulty, e.g., fine-grained visual classification in the context of SSL (SS-FGVC). The increased recognition difficulty on fine-grained unlabeled data spells disaster for pseudo-labeling accuracy, resulting in poor performance of the SSL model. To tackle this challenge, we propose Soft Label Selection with Confidence-Aware Clustering based on Class Transition Tracking (SoC) by reconstructing the pseudo-label selection process by jointly optimizing Expansion Objective and Shrinkage Objective, which is based on a soft label manner. Respectively, the former objective encourages soft labels to absorb more candidate classes to ensure the attendance of ground-truth class, while the latter encourages soft labels to reject more noisy classes, which is theoretically proved to be equivalent to entropy minimization. In comparisons with various state-of-the-art methods, our approach demonstrates its superior performance in SS-FGVC. Checkpoints and source code are available at https://github.com/NJUyued/SoC4SS-FGVC.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

START: A Generalized State Space Model with Saliency-Driven Token-Aware Transformation

Jintao Guo
Lei Qi
Yinghuan Shi
Yang Gao

Domain Generalization (DG) aims to enable models to generalize to unseen target domains by learning from multiple source domains. Existing DG methods primarily rely on convolutional neural networks (CNNs), which inherently learn texture biases due to their limited receptive fields, making them prone to overfitting source domains. While some works have introduced transformer-based methods (ViTs) for DG to leverage the global receptive field, these methods incur high computational costs due to the quadratic complexity of self-attention. Recently, advanced state space models (SSMs), represented by Mamba, have shown promising results in supervised learning tasks by achieving linear complexity in sequence length during training and fast RNN-like computation during inference. Inspired by this, we investigate the generalization ability of the Mamba model under domain shifts and find that input-dependent matrices within SSMs could accumulate and amplify domain-specific features, thus hindering model generalization. To address this issue, we propose a novel SSM-based architecture with saliency-based token-aware transformation (namely START), which achieves state-of-the-art (SOTA) performances and offers a competitive alternative to CNNs and ViTs. Our START can selectively perturb and suppress domain-specific features in salient tokens within the input-dependent matrices of SSMs, thus effectively reducing the discrepancy between different domains. Extensive experiments on five benchmarks demonstrate that START outperforms existing SOTA DG methods with efficient linear complexity. Our code is available at https: //github. com/lingeringlight/START.

PDF Details DOI

NeurIPS Conference 2023 Conference Paper

FreeMask: Synthetic Images with Dense Annotations Make Stronger Segmentation Models

Lihe Yang
Xiaogang Xu
Bingyi Kang
Yinghuan Shi
Hengshuang Zhao

Semantic segmentation has witnessed tremendous progress due to the proposal of various advanced network architectures. However, they are extremely hungry for delicate annotations to train, and the acquisition is laborious and unaffordable. Therefore, we present FreeMask in this work, which resorts to synthetic images from generative models to ease the burden of both data collection and annotation procedures. Concretely, we first synthesize abundant training images conditioned on the semantic masks provided by realistic datasets. This yields extra well-aligned image-mask training pairs for semantic segmentation models. We surprisingly observe that, solely trained with synthetic images, we already achieve comparable performance with real ones (e. g. , 48. 3 vs. 48. 5 mIoU on ADE20K, and 49. 3 vs. 50. 5 on COCO-Stuff). Then, we investigate the role of synthetic images by joint training with real images, or pre-training for real images. Meantime, we design a robust filtering principle to suppress incorrectly synthesized regions. In addition, we propose to inequally treat different semantic masks to prioritize those harder ones and sample more corresponding synthetic images for them. As a result, either jointly trained or pre-trained with our filtered and re-sampled synthesized images, segmentation models can be greatly enhanced, e. g. , from 48. 7 to 52. 0 on ADE20K.

PDF Details

NeurIPS Conference 2022 Conference Paper

Improving Barely Supervised Learning by Discriminating Unlabeled Samples with Super-Class

Guan Gui
Zhen Zhao
Lei Qi
Luping Zhou
Lei Wang
Yinghuan Shi

In semi-supervised learning (SSL), a common practice is to learn consistent information from unlabeled data and discriminative information from labeled data to ensure both the immutability and the separability of the classification model. Existing SSL methods suffer from failures in barely-supervised learning (BSL), where only one or two labels per class are available, as the insufficient labels cause the discriminative information being difficult or even infeasible to learn. To bridge this gap, we investigate a simple yet effective way to leverage unlabeled samples for discriminative learning, and propose a novel discriminative information learning module to benefit model training. Specifically, we formulate the learning objective of discriminative information at the super-class level and dynamically assign different classes into different super-classes based on model performance improvement. On top of this on-the-fly process, we further propose a distribution-based loss to learn discriminative information by utilizing the similarity relationship between samples and super-classes. It encourages the unlabeled samples to stay closer to the distribution of their corresponding super-class than those of others. Such a constraint is softer than the direct assignment of pseudo labels, while the latter could be very noisy in BSL. We compare our method with state-of-the-art SSL and BSL methods through extensive experiments on standard SSL benchmarks. Our method can achieve superior results, \eg, an average accuracy of 76. 76\% on CIFAR-10 with merely 1 label per class.

PDF Details

AAAI Conference 2022 Conference Paper

LaSSL: Label-Guided Self-Training for Semi-supervised Learning

Zhen Zhao
Luping Zhou
Lei Wang
Yinghuan Shi
Yang Gao

The key to semi-supervised learning (SSL) is to explore adequate information to leverage the unlabeled data. Current dominant approaches aim to generate pseudolabels on weakly augmented instances and train models on their corresponding strongly augmented variants with high-confidence results. However, such methods are limited in excluding samples with low-confidence pseudo-labels and under-utilization of the label information. In this paper, we emphasize the cruciality of the label information and propose a Label-guided Self-training approach to Semi-supervised Learning (LaSSL), which improves pseudo-label generations from two mutually boosted strategies. First, with the ground-truth labels and iteratively-polished pseudolabels, we explore instance relations among all samples and then minimize a class-aware contrastive loss to learn discriminative feature representations that make same-class samples gathered and different-class samples scattered. Second, on top of improved feature representations, we propagate the label information to the unlabeled samples across the potential data manifold at the feature-embedding level, which can further improve the labelling of samples with reference to their neighbours. These two strategies are seamlessly integrated and mutually promoted across the whole training process. We evaluate LaSSL on several classification benchmarks under partially labeled settings and demonstrate its superiority over the state-of-the-art approaches.

PDF Details

AIIM Journal 2021 Journal Article

Interactive medical image segmentation via a point-based interaction

Jian Zhang
Yinghuan Shi
Jinquan Sun
Lei Wang
Luping Zhou
Yang Gao
Dinggang Shen

Due to low tissue contrast, irregular shape, and large location variance, segmenting the objects from different medical imaging modalities (e. g. , CT, MR) is considered as an important yet challenging task. In this paper, a novel method is presented for interactive medical image segmentation with the following merits. (1) Its design is fundamentally different from previous pure patch-based and image-based segmentation methods. It is observed that during delineation, the physician repeatedly check the intensity from area inside-object to outside-object to determine the boundary, which indicates that comparison in an inside-out manner is extremely important. Thus, the method innovatively models the segmentation task as learning the representation of bi-directional sequential patches, starting from (or ending in) the given central point of the object. This can be realized by the proposed ConvRNN network embedded with a gated memory propagation unit. (2) Unlike previous interactive methods (requiring bounding box or seed points), the proposed method only asks the physician to merely click on the rough central point of the object before segmentation, which could simultaneously enhance the performance and reduce the segmentation time. (3) The method is utilized in a multi-level framework for better performance. It has been systematically evaluated in three different segmentation tasks, including CT kidney tumor, MR prostate, and PROMISE12 challenge, showing promising results compared with state-of-the-art methods.

Details DOI

JBHI Journal 2021 Journal Article

Learning-Based Computer-Aided Prescription Model for Parkinson's Disease: A Data-Driven Perspective

Yinghuan Shi
Wanqi Yang
Kim-Han Thung
Hao Wang
Yang Gao
Yang Pan
Li Zhang
Dinggang Shen

In this article, we study a novel problem: “automatic prescription recommendation for PD patients. ” To realize this goal, we first build a dataset by collecting 1) symptoms of PD patients, and 2) their prescription drug provided by neurologists. Then, we build a novel computer-aided prescription model by learning the relation between observed symptoms and prescription drug. Finally, for the new coming patients, we could recommend (predict) suitable prescription drug on their observed symptoms by our prescription model. From the methodology part, our proposed model, namely Prescription viA Learning lAtent Symptoms (PALAS), could recommend prescription using the multi-modality representation of the data. In PALAS, a latent symptom space is learned to better model the relationship between symptoms and prescription drug, as there is a large semantic gap between them. Moreover, we present an efficient alternating optimization method for PALAS. We evaluated our method using the data collected from 136 PD patients at Nanjing Brain Hospital, which can be regarded as a large dataset in PD research community. The experimental results demonstrate the effectiveness and clinical potential of our method in this recommendation task, if compared with other competing methods.

Details DOI

YNIMG Journal 2021 Journal Article

Recycling diagnostic MRI for empowering brain morphometric research – Critical & practical assessment on learning-based image super-resolution

Gaoping Liu
Zehong Cao
Qiang Xu
Qirui Zhang
Fang Yang
Xinyu Xie
Jingru Hao
Yinghuan Shi

Preliminary studies have shown the feasibility of deep learning (DL)-based super-resolution (SR) technique for reconstructing thick-slice/gap diagnostic MR images into high-resolution isotropic data, which would be of great significance for brain research field if the vast amount of diagnostic MRI data could be successively put into brain morphometric study. However, less evidence has addressed the practicability of the strategy, because lack of a large-sample available real data for constructing DL model. In this work, we employed a large cohort (n = 2052) of peculiar data with both low through-plane resolution diagnostic and high-resolution isotropic brain MR images from identical subjects. By leveraging a series of SR approaches, including a proposed novel DL algorithm of Structure Constrained Super Resolution Network (SCSRN), the diagnostic images were transformed to high-resolution isotropic data to meet the criteria of brain research in voxel-based and surface-based morphometric analyses. We comprehensively assessed image quality and the practicability of the reconstructed data in a variety of morphometric analysis scenarios. We further compared the performance of SR approaches to the ground truth high-resolution isotropic data. The results showed (i) DL-based SR algorithms generally improve the quality of diagnostic images and render morphometric analysis more accurate, especially, with the most superior performance of the novel approach of SCSRN. (ii) Accuracies vary across brain structures and methods, and (iii) performance increases were higher for voxel than for surface based approaches. This study supports that DL-based image super-resolution potentially recycle huge amount of routine diagnostic brain MRI deposited in sleeping state, and turning them into useful data for neurometric research.

Details DOI

JBHI Journal 2020 Journal Article

An Effective MR-Guided CT Network Training for Segmenting Prostate in CT Images

Wanqi Yang
Yinghuan Shi
Sang Hyun Park
Ming Yang
Yang Gao
Dinggang Shen

Segmentation of prostate in medical imaging data (e. g. , CT, MRI, TRUS) is often considered as a critical yet challenging task for radiotherapy treatment. It is relatively easier to segment prostate from MR images than from CT images, due to better soft tissue contrast of the MR images. For segmenting prostate from CT images, most previous methods mainly used CT alone, and thus their performances are often limited by low tissue contrast in the CT images. In this article, we explore the possibility of using indirect guidance from MR images for improving prostate segmentation in the CT images. In particular, we propose a novel deep transfer learning approach, i. e. , MR-guided CT network training (namely MICS-NET), which can employ MR images to help better learning of features in CT images for prostate segmentation. In MICS-NET, the guidance from MRI consists of two steps: (1) learning informative and transferable features from MRI and then transferring them to CT images in a cascade manner, and (2) adaptively transferring the prostate likelihood of MRI model (i. e. , well-trained convnet by purely using MR images) with a view consistency constraint. To illustrate the effectiveness of our approach, we evaluate MICS-NET on a real CT prostate image set, with the manual delineations available as the ground truth for evaluation. Our methods generate promising segmentation results which achieve (1) six percentages higher Dice Ratio than the CT model purely using CT images and (2) comparable performance with the MRI model purely using MR images.

Details DOI

IJCAI Conference 2020 Conference Paper

Asymmetric Distribution Measure for Few-shot Learning

Wenbin Li
Lei Wang
Jing Huo
Yinghuan Shi
Yang Gao
Jiebo Luo

The core idea of metric-based few-shot image classification is to directly measure the relations between query images and support classes to learn transferable feature embeddings. Previous work mainly focuses on image-level feature representations, which actually cannot effectively estimate a class's distribution due to the scarcity of samples. Some recent work shows that local descriptor based representations can achieve richer representations than image-level based representations. However, such works are still based on a less effective instance-level metric, especially a symmetric metric, to measure the relation between a query image and a support class. Given the natural asymmetric relation between a query image and a support class, we argue that an asymmetric measure is more suitable for metric-based few-shot learning. To that end, we propose a novel Asymmetric Distribution Measure (ADM) network for few-shot learning by calculating a joint local and global asymmetric measure between two multivariate local distributions of a query and a class. Moreover, a task-aware Contrastive Measure Strategy (CMS) is proposed to further enhance the measure function. On popular miniImageNet and tieredImageNet, ADM can achieve the state-of-the-art results, validating our innovative design of asymmetric distribution measures for few-shot learning. The source code can be downloaded from https: //github. com/WenbinLee/ADM. git.

PDF Details DOI

IJCAI Conference 2020 Conference Paper

Consistent MetaReg: Alleviating Intra-task Discrepancy for Better Meta-knowledge

Pinzhuo Tian
Lei Qi
Shaokang Dong
Yinghuan Shi
Yang Gao

In the few-shot learning scenario, the data-distribution discrepancy between training data and test data in a task usually exists due to the limited data. However, most existing meta-learning approaches seldom consider this intra-task discrepancy in the meta-training phase which might deteriorate the performance. To overcome this limitation, we develop a new consistent meta-regularization method to reduce the intra-task data-distribution discrepancy. Moreover, the proposed meta-regularization method could be readily inserted into existing optimization-based meta-learning models to learn better meta-knowledge. Particularly, we provide the theoretical analysis to prove that using the proposed meta-regularization, the conventional gradient-based meta-learning method can reach the lower regret bound. The extensive experiments also demonstrate the effectiveness of our method, which indeed improves the performances of the state-of-the-art gradient-based meta-learning models in the few-shot classification task.

PDF Details DOI

AAAI Conference 2020 Conference Paper

Differentiable Meta-Learning Model for Few-Shot Semantic Segmentation

Pinzhuo Tian
Zhangkai Wu
Lei Qi
Lei Wang
Yinghuan Shi
Yang Gao

To address the annotation scarcity issue in some cases of semantic segmentation, there have been a few attempts to develop the segmentation model in the few-shot learning paradigm. However, most existing methods only focus on the traditional 1-way segmentation setting (i. e. , one image only contains a single object). This is far away from practical semantic segmentation tasks where the K-way setting (K >1) is usually required by performing the accurate multi-object segmentation. To deal with this issue, we formulate the fewshot semantic segmentation task as a learning-based pixel classiﬁcation problem, and propose a novel framework called MetaSegNet based on meta-learning. In MetaSegNet, an architecture of embedding module consisting of the global and local feature branches is developed to extract the appropriate meta-knowledge for the few-shot segmentation. Moreover, we incorporate a linear model into MetaSegNet as a base learner to directly predict the label of each pixel for the multiobject segmentation. Furthermore, our MetaSegNet can be trained by the episodic training mechanism in an end-to-end manner from scratch. Experiments on two popular semantic segmentation datasets, i. e. , PASCAL VOC and COCO, reveal the effectiveness of the proposed MetaSegNet in the K-way few-shot semantic segmentation task.

PDF Details

AAAI Conference 2017 Conference Paper

Beyond IID: Learning to Combine Non-IID Metrics for Vision Tasks

Yinghuan Shi
Wenbin Li
Yang Gao
Longbing Cao
Dinggang Shen

Metric learning has been widely employed, especially in various computer vision tasks, with the fundamental assumption that all samples (e. g. , regions/superpixels in images/videos) are independent and identically distributed (IID). However, since the samples are usually spatially-connected or temporally-correlated with their physically-connected neighbours, they are not IID (non-IID for short), which cannot be directly handled by existing methods. Thus, we propose to learn and integrate non-IID metrics (NIME). To incorporate the non-IID spatial/temporal relations, instead of directly using non-IID features and metric learning as previous methods, NIME ﬁrst builds several non-IID representations on original (non-IID) features by various graph kernel functions, and then automatically learns the metric under the best combination of various non-IID representations. NIME is applied to solve two typical computer vision tasks: interactive image segmentation and histology image identiﬁcation. The results show that learning and integrating non-IID metrics improves the performance, compared to the IID methods. Moreover, our method achieves results comparable or better than that of the state-of-the-arts.

PDF Details