Can Peng

Can Peng, Kun Zhao, Tianren Wang et al.

The continual appearance of new objects in the visual world poses considerable challenges for current deep learning methods in real-world deployments. The challenge of new task learning is often exacerbated by the scarcity of data for the new categories due to rarity or cost. Here we explore the important task of Few-Shot Class-Incremental Learning (FSCIL) and its extreme data scarcity condition of one-shot. An ideal FSCIL model needs to perform well on all classes, regardless of their presentation order or paucity of data. It also needs to be robust to open-set real-world conditions and be easily adapted to the new tasks that always arise in the field. In this paper, we first reevaluate the current task setting and propose a more comprehensive and practical setting for the FSCIL task. Then, inspired by the similarity of the goals for FSCIL and modern face recognition systems, we propose our method -- Augmented Angular Loss Incremental Classification or ALICE. In ALICE, instead of the commonly used cross-entropy loss, we propose to use the angular penalty loss to obtain well-clustered features. As the obtained features not only need to be compactly clustered but also diverse enough to maintain generalization for future incremental classes, we further discuss how class augmentation, data augmentation, and data balancing affect classification performance. Experiments on benchmark datasets, including CIFAR100, miniImageNet, and CUB200, demonstrate the improved performance of ALICE over the state-of-the-art FSCIL methods.

Meng Li, Chaoyi Li, Can Peng et al.

Deep learning techniques have become widely utilized in histopathology image classification due to their superior performance. However, this success heavily relies on the availability of substantial labeled data, which necessitates extensive and costly manual annotation by domain experts. To address this challenge, researchers have recently employed generative models to synthesize data for augmentation, thereby enhancing classification model performance. Traditionally, this involves generating synthetic data first and then training the classification model with both synthetic and real data, which creates a two-stage, time-consuming workflow. To overcome this limitation, we propose an innovative unified framework that integrates the data generation and model training stages into a unified process. Our approach utilizes a pure Vision Transformer (ViT)-based conditional Generative Adversarial Network (cGAN) model to simultaneously handle both image synthesis and classification. An additional classification head is incorporated into the cGAN model to enable simultaneous classification of histopathology images. To improve training stability and enhance the quality of generated data, we introduce a conditional class projection technique that helps maintain class separation during the generation process. We also employ a dynamic multi-loss weighting mechanism to effectively balance the losses of the classification tasks. Furthermore, our selective augmentation mechanism actively selects the most suitable generated images for data augmentation to further improve performance. Extensive experiments on histopathology datasets show that our unified synthetic augmentation framework consistently enhances the performance of histopathology image classification models.

Yuyuan Liu, Yiping Ji, Anjie Le et al.

Finetuning Large Vision-Language Models with reinforcement learning has emerged as a promising approach to enhance their capability in object-level grounding. However, existing methods, mainly based on GRPO, assign rewards at the response level. Such sparse reward, often criterion-induced, leads to minimal learning signals when all candidate responses fail in challenging scenarios. In this work, we propose a group-revision optimisation paradigm that enhances learning on hard cases. It begins with a sampled initial response and generates a set of revised candidates to explore improved grounding outcomes. Inspired by reward shaping, we introduce a consolidation process that quantifies each candidate's improvement over the initial attempt and converts it into informative shaping signals. These signals are used to both refine the reward and modulate the advantage, amplifying the influence of high-quality revisions. Our method achieves consistent gains across referring and reasoning segmentation, REC, and counting benchmarks compared with prior GRPO-based models. Our code is available at https://github.com/yyliu01/GroupRevision.

Yingyu Yang, Qianye Yang, Can Peng et al.

Fetal echocardiography is essential for detecting congenital heart disease (CHD), facilitating pregnancy management, optimized delivery planning, and timely postnatal interventions. Among standard imaging planes, the four-chamber (4CH) view provides comprehensive information for CHD diagnosis, where clinicians carefully inspect the end-diastolic (ED) and end-systolic (ES) phases to evaluate cardiac structure and motion. Automated detection of these cardiac phases is thus a critical component toward fully automated CHD analysis. Yet, in the absence of fetal electrocardiography (ECG), manual identification of ED and ES frames remains a labor-intensive bottleneck. We present ORBIT (Orientation-Robust Beat Inference from Trajectories), a self-supervised framework that identifies cardiac phases without manual annotations under various fetal heart orientation. ORBIT employs registration as self-supervision task and learns a latent motion trajectory of cardiac deformation, whose turning points capture transitions between cardiac relaxation and contraction, enabling accurate and orientation-robust localization of ED and ES frames across diverse fetal positions. Trained exclusively on normal fetal echocardiography videos, ORBIT achieves consistent performance on both normal (MAE = 1.9 frames for ED and 1.6 for ES) and CHD cases (MAE = 2.4 frames for ED and 2.1 for ES), outperforming existing annotation-free approaches constrained by fixed orientation assumptions. These results highlight the potential of ORBIT to facilitate robust cardiac phase detection directly from 4CH fetal echocardiography.

Can Peng, Piotr Koniusz, Kaiyu Guo et al.

Deep learning models suffer from catastrophic forgetting when being fine-tuned with samples of new classes. This issue becomes even more pronounced when faced with the domain shift between training and testing data. In this paper, we study the critical and less explored Domain-Generalized Class-Incremental Learning (DGCIL). We design a DGCIL approach that remembers old classes, adapts to new classes, and can classify reliably objects from unseen domains. Specifically, our loss formulation maintains classification boundaries and suppresses the domain-specific information of each class. With no old exemplars stored, we use knowledge distillation and estimate old class prototype drift as incremental training advances. Our prototype representations are based on multivariate Normal distributions whose means and covariances are constantly adapted to changing model features to represent old classes well by adapting to the feature space drift. For old classes, we sample pseudo-features from the adapted Normal distributions with the help of Cholesky decomposition. In contrast to previous pseudo-feature sampling strategies that rely solely on average mean prototypes, our method excels at capturing varying semantic information. Experiments on several benchmarks validate our claims.

Yuyuan Liu, Can Peng, Yingyu Yang et al.

Recent progress in deep learning has significantly advanced CT image analysis, particularly for segmentation tasks. However, these advances are largely confined to image-level pattern recognition, with most methods lacking explicit anatomical or contextual reasoning. Large vision-language models introduce linguistic context into image analysis, yet most approaches typically focus on a single task, which is insufficient for clinical workflow analysis that requires multiple fine-grained types of analysis, such as anatomy detection and segmentation. In this paper, we propose a unified autoregressive framework that integrates language-guided visual reasoning into CT interpretation. Our method introduces task-routing tokens that trigger detection and segmentation heads conditioned on the hidden states of a large vision-language model, enabling coherent generation of visual outputs (e.g., masks and bounding boxes) and textual reasonings. To progressively enhance localisation accuracy and semantic clarity, we further design a "closer-look" mechanism that allows the model to perform progressive coarse-to-fine visits to regions of interest under refined fields of view. To support model training and evaluation, we curated a new multimodal CT dataset containing pixel-wise masks, bounding boxes, spatial prompts, and structured descriptions for visual objects constructed through an AI-assisted annotation process with human verification. Experiments on public benchmarks demonstrate consistent improvements over the SoTA, achieving up to 1.0% Dice on BTCV and 1.7% Dice on MosMed+, while additionally providing appearance reasoning outputs. The code and dataset will be available.

Yingyu Yang, Qianye Yang, Kangning Cui et al.

The identification of cardiac phase is an essential step for analysis and diagnosis of cardiac function. Automatic methods, especially data-driven methods for cardiac phase detection, typically require extensive annotations, which is time-consuming and labor-intensive. In this paper, we present an unsupervised framework for end-diastole (ED) and end-systole (ES) detection through self-supervised learning of latent cardiac motion trajectories from 4-chamber-view echocardiography videos. Our method eliminates the need for manual annotations, including ED and ES indices, segmentation, or volumetric measurements, by training a reconstruction model to encode interpretable spatiotemporal motion patterns. Evaluated on the EchoNet-Dynamic benchmark, the approach achieves mean absolute error (MAE) of 3 frames (58.3 ms) for ED and 2 frames (38.8 ms) for ES detection, matching state-of-the-art supervised methods. Extended to fetal echocardiography, the model demonstrates robust performance with MAE 1.46 frames (20.7 ms) for ED and 1.74 frames (25.3 ms) for ES, despite the fact that the fetal heart model is built using non-standardized heart views due to fetal heart positioning variability. Our results demonstrate the potential of the proposed latent motion trajectory strategy for cardiac phase detection in adult and fetal echocardiography. This work advances unsupervised cardiac motion analysis, offering a scalable solution for clinical populations lacking annotated data. Code will be released at https://github.com/YingyuYyy/CardiacPhase.

Yuyuan Liu, Yuanhong Chen, Chong Wang et al.

Segment Anything Model 2 (SAM2) exhibits strong generalisation for promptable segmentation in video clips; however, its integration with the audio modality remains underexplored. Existing approaches mainly follow two directions: (1) injecting adapters into the image encoder to receive audio signals, which incurs efficiency costs during prompt engineering, and (2) leveraging additional foundation models to generate visual prompts for the sounding objects, which are often imprecisely localised, leading to misguidance in SAM2. Moreover, these methods overlook the rich semantic interplay between hierarchical visual features and other modalities, resulting in suboptimal cross-modal fusion. In this work, we propose AuralSAM2, comprising the novel AuralFuser module, which externally attaches to SAM2 to integrate features from different modalities and generate feature-level prompts, guiding SAM2's decoder in segmenting sounding targets. Such integration is facilitated by a feature pyramid, further refining semantic understanding and enhancing object awareness in multimodal scenarios. Additionally, the audio-guided contrastive learning is introduced to explicitly align audio and visual representations and to also mitigate biases caused by dominant visual patterns. Results on public benchmarks show that our approach achieves remarkable improvements over the previous methods in the field. Code is available at https://github.com/yyliu01/AuralSAM2.

Mingjie Wu, Chenggui Yang, Huihua Wang et al.

The UAV technology is gradually maturing and can provide extremely powerful support for smart agriculture and precise monitoring. Currently, there is no dataset related to green walnuts in the field of agricultural computer vision. Thus, in order to promote the algorithm design in the field of agricultural computer vision, we used UAV to collect remote-sensing data from 8 walnut sample plots. Considering that green walnuts are subject to various lighting conditions and occlusion, we constructed a large-scale dataset with a higher-granularity of target features - WalnutData. This dataset contains a total of 30,240 images and 706,208 instances, and there are 4 target categories: being illuminated by frontal light and unoccluded (A1), being backlit and unoccluded (A2), being illuminated by frontal light and occluded (B1), and being backlit and occluded (B2). Subsequently, we evaluated many mainstream algorithms on WalnutData and used these evaluation results as the baseline standard. The dataset and all evaluation results can be obtained at https://github.com/1wuming/WalnutData.

Yayong Li, Peyman Moghadam, Can Peng et al.

Node classification with Graph Neural Networks (GNN) under a fixed set of labels is well known in contrast to Graph Few-Shot Class Incremental Learning (GFSCIL), which involves learning a GNN classifier as graph nodes and classes growing over time sporadically. We introduce inductive GFSCIL that continually learns novel classes with newly emerging nodes while maintaining performance on old classes without accessing previous data. This addresses the practical concern of transductive GFSCIL, which requires storing the entire graph with historical data. Compared to the transductive GFSCIL, the inductive setting exacerbates catastrophic forgetting due to inaccessible previous data during incremental training, in addition to overfitting issue caused by label sparsity. Thus, we propose a novel method, called Topology-based class Augmentation and Prototype calibration (TAP). To be specific, it first creates a triple-branch multi-topology class augmentation method to enhance model generalization ability. As each incremental session receives a disjoint subgraph with nodes of novel classes, the multi-topology class augmentation method helps replicate such a setting in the base session to boost backbone versatility. In incremental learning, given the limited number of novel class samples, we propose an iterative prototype calibration to improve the separation of class prototypes. Furthermore, as backbone fine-tuning poses the feature distribution drift, prototypes of old classes start failing over time, we propose the prototype shift method for old classes to compensate for the drift. We showcase the proposed method on four datasets.

Pramit Saha, Felix Wagner, Divyanshu Mishra et al.

Effective training of large Vision-Language Models (VLMs) on resource-constrained client devices in Federated Learning (FL) requires the usage of parameter-efficient fine-tuning (PEFT) strategies. To this end, we demonstrate the impact of two factors \textit{viz.}, client-specific layer importance score that selects the most important VLM layers for fine-tuning and inter-client layer diversity score that encourages diverse layer selection across clients for optimal VLM layer selection. We first theoretically motivate and leverage the principal eigenvalue magnitude of layerwise Neural Tangent Kernels and show its effectiveness as client-specific layer importance score. Next, we propose a novel layer updating strategy dubbed F$^3$OCUS that jointly optimizes the layer importance and diversity factors by employing a data-free, multi-objective, meta-heuristic optimization on the server. We explore 5 different meta-heuristic algorithms and compare their effectiveness for selecting model layers and adapter layers towards PEFT-FL. Furthermore, we release a new MedVQA-FL dataset involving overall 707,962 VQA triplets and 9 modality-specific clients and utilize it to train and evaluate our method. Overall, we conduct more than 10,000 client-level experiments on 6 Vision-Language FL task settings involving 58 medical image datasets and 4 different VLM architectures of varying sizes to demonstrate the effectiveness of the proposed method.

Anjie Le, Can Peng, Yuyuan Liu et al.

In computer vision, machine unlearning aims to remove the influence of specific visual concepts or training images without retraining from scratch. Studies show that existing approaches often modify the classifier while leaving internal representations intact, resulting in incomplete forgetting. In this work, we extend the notion of unlearning to the representation level, deriving a three-term interplay between forgetting efficacy, retention fidelity, and class separation. Building on Neural Collapse theory, we show that the orthogonal projection of a simplex Equiangular Tight Frame (ETF) remains an ETF in a lower dimensional space, yielding a provably optimal forgetting operator. We further introduce the Representation Unlearning Score (RUS) to quantify representation-level forgetting and retention fidelity. Building on this, we introduce POUR (Provably Optimal Unlearning of Representations), a geometric projection method with closed-form (POUR-P) and a feature-level unlearning variant under a distillation scheme (POUR-D). Experiments on CIFAR-10/100 and PathMNIST demonstrate that POUR achieves effective unlearning while preserving retained knowledge, outperforming state-of-the-art unlearning methods on both classification-level and representation-level metrics.

Can Peng, Yuyuan Liu, Yingyu Yang et al.

Federated Learning (FL) enables collaborative model training across distributed clients while preserving data privacy. However, the performance of deep learning often deteriorates in FL due to decentralized and heterogeneous data. This challenge is further amplified in multi-label scenarios, where data exhibit complex characteristics such as label co-occurrence, inter-label dependency, and discrepancies between local and global label relationships. While most existing FL research primarily focuses on single-label classification, many real-world applications, particularly in domains such as medical imaging, often involve multi-label settings. In this paper, we address this important yet underexplored scenario in FL, where clients hold multi-label data with skewed label distributions. Neural Collapse (NC) describes a geometric structure in the latent feature space where features of each class collapse to their class mean with vanishing intra-class variance, and the class means form a maximally separated configuration. Motivated by this theory, we propose a method to align feature distributions across clients and to learn high-quality, well-clustered representations. To make the NC-structure applicable to multi-label settings, where image-level features may contain multiple semantic concepts, we introduce a feature disentanglement module that extracts semantically specific features. The clustering of these disentangled class-wise features is guided by a predefined shared NC structure, which mitigates potential conflicts between client models due to diverse local data distributions. In addition, we design regularisation losses to encourage compact clustering in the latent feature space. Experiments conducted on four benchmark datasets across eight diverse settings demonstrate that our approach outperforms existing methods, validating its effectiveness in this challenging FL scenario.

Can Peng, Qianhui Men, Pramit Saha et al.

Federated learning seeks to foster collaboration among distributed clients while preserving the privacy of their local data. Traditionally, federated learning methods assume a fixed setting in which client data and learning objectives remain constant. However, in real-world scenarios, new clients may join, and existing clients may expand the segmentation label set as task requirements evolve. In such a dynamic federated analysis setup, the conventional federated communication strategy of model aggregation per communication round is suboptimal. As new clients join, this strategy requires retraining, linearly increasing communication and computation overhead. It also imposes requirements for synchronized communication, which is difficult to achieve among distributed clients. In this paper, we propose a federated continual learning strategy that employs a one-time model aggregation at the server through multi-model distillation. This approach builds and updates the global model while eliminating the need for frequent server communication. When integrating new data streams or onboarding new clients, this approach efficiently reuses previous client models, avoiding the need to retrain the global model across the entire federation. By minimizing communication load and bypassing the need to put unchanged clients online, our approach relaxes synchronization requirements among clients, providing an efficient and scalable federated analysis framework suited for real-world applications. Using multi-class 3D abdominal CT segmentation as an application task, we demonstrate the effectiveness of the proposed approach.

Tianren Wang, Can Peng, Teng Zhang et al.

With the excellent disentanglement properties of state-of-the-art generative models, image editing has been the dominant approach to control the attributes of synthesised face images. However, these edited results often suffer from artifacts or incorrect feature rendering, especially when there is a large discrepancy between the image to be edited and the desired feature set. Therefore, we propose a new approach to mapping the latent vectors of the generative model to the scaling factors through solving a set of multivariate linear equations. The coefficients of the equations are the eigenvectors of the weight parameters of the pre-trained model, which form the basis of a hyper coordinate system. The qualitative and quantitative results both show that the proposed method outperforms the baseline in terms of image diversity. In addition, the method is much more time-efficient because you can obtain synthesised images with desirable features directly from the latent vectors, rather than the former process of editing randomly generated images requiring many processing steps.

Can Peng, Kun Zhao, Sam Maksoud et al.

To accommodate rapid changes in the real world, the cognition system of humans is capable of continually learning concepts. On the contrary, conventional deep learning models lack this capability of preserving previously learned knowledge. When a neural network is fine-tuned to learn new tasks, its performance on previously trained tasks will significantly deteriorate. Many recent works on incremental object detection tackle this problem by introducing advanced regularization. Although these methods have shown promising results, the benefits are often short-lived after the first incremental step. Under multi-step incremental learning, the trade-off between old knowledge preserving and new task learning becomes progressively more severe. Thus, the performance of regularization-based incremental object detectors gradually decays for subsequent learning steps. In this paper, we aim to alleviate this performance decay on multi-step incremental detection tasks by proposing a dilatable incremental object detector (DIODE). For the task-shared parameters, our method adaptively penalizes the changes of important weights for previous tasks. At the same time, the structure of the model is dilated or expanded by a limited number of task-specific parameters to promote new task learning. Extensive experiments on PASCAL VOC and COCO datasets demonstrate substantial improvements over the state-of-the-art methods. Notably, compared with the state-of-the-art methods, our method achieves up to 6.0% performance improvement by increasing the number of parameters by just 1.2% for each newly learned task.

Sam Maksoud, Kun Zhao, Can Peng et al.

This paper addresses the scalability problem of Bayesian deep neural networks. The performance of deep neural networks is undermined by the fact that these algorithms have poorly calibrated measures of uncertainty. This restricts their application in high risk domains such as computer aided diagnosis and autonomous vehicle navigation. Bayesian Deep Learning (BDL) offers a promising method for representing uncertainty in neural network. However, BDL requires a separate set of parameters to store the mean and standard deviation of model weights to learn a distribution. This results in a prohibitive 2-fold increase in the number of model parameters. To address this problem we present a method for performing BDL, namely Kernel Seed Networks (KSN), which does not require a 2-fold increase in the number of parameters. KSNs use 1x1 Convolution operations to learn a compressed latent space representation of the parameter distribution. In this paper we show how this allows KSNs to outperform conventional BDL methods while reducing the number of required parameters by up to a factor of 6.6.

Can Peng, Kun Zhao, Sam Maksoud et al.

Incremental learning requires a model to continually learn new tasks from streaming data. However, traditional fine-tuning of a well-trained deep neural network on a new task will dramatically degrade performance on the old task -- a problem known as catastrophic forgetting. In this paper, we address this issue in the context of anchor-free object detection, which is a new trend in computer vision as it is simple, fast, and flexible. Simply adapting current incremental learning strategies fails on these anchor-free detectors due to lack of consideration of their specific model structures. To deal with the challenges of incremental learning on anchor-free object detectors, we propose a novel incremental learning paradigm called Selective and Inter-related Distillation (SID). In addition, a novel evaluation metric is proposed to better assess the performance of detectors under incremental learning conditions. By selective distilling at the proper locations and further transferring additional instance relation knowledge, our method demonstrates significant advantages on the benchmark datasets PASCAL VOC and COCO.

Can Peng, Kun Zhao, Brian C. Lovell

The human vision and perception system is inherently incremental where new knowledge is continually learned over time whilst existing knowledge is retained. On the other hand, deep learning networks are ill-equipped for incremental learning. When a well-trained network is adapted to new categories, its performance on the old categories will dramatically degrade. To address this problem, incremental learning methods have been explored which preserve the old knowledge of deep learning models. However, the state-of-the-art incremental object detector employs an external fixed region proposal method that increases overall computation time and reduces accuracy comparing to Region Proposal Network (RPN) based object detectors such as Faster RCNN. The purpose of this paper is to design an efficient end-to-end incremental object detector using knowledge distillation. We first evaluate and analyze the performance of the RPN-based detector with classic distillation on incremental detection tasks. Then, we introduce multi-network adaptive distillation that properly retains knowledge from the old categories when fine-tuning the model for new task. Experiments on the benchmark datasets, PASCAL VOC and COCO, demonstrate that the proposed incremental detector based on Faster RCNN is more accurate as well as being 13 times faster than the baseline detector.

Can Peng, Kun Zhao, Arnold Wiliem et al.

The condition of the Glomeruli, or filter sacks, in renal Direct Immunofluorescence (DIF) specimens is a critical indicator for diagnosing kidney diseases. A digital pathology system which digitizes a glass histology slide into a Whole Slide Image (WSI) and then automatically detects and zooms in on the glomeruli with a higher magnification objective will be extremely helpful for pathologists. In this paper, using glomerulus detection as the study case, we provide analysis and observations on several important issues to help with the development of Computer Aided Diagnostic (CAD) systems to process WSIs. Large image resolution, large file size, and data scarcity are always challenging to deal with. To this end, we first examine image downsampling rates in terms of their effect on detection accuracy. Second, we examine the impact of image compression. Third, we examine the relationship between the size of the training set and detection accuracy. To understand the above issues, experiments are performed on the state-of-the-art detectors: Faster R-CNN, R-FCN, Mask R-CNN and SSD. Critical findings are observed: (1) The best balance between detection accuracy, detection speed and file size is achieved at 8 times downsampling captured with a $40\times$ objective; (2) compression which reduces the file size dramatically, does not necessarily have an adverse effect on overall accuracy; (3) reducing the amount of training data to some extents causes a drop in precision but has a negligible impact on the recall; (4) in most cases, Faster R-CNN achieves the best accuracy in the glomerulus detection task. We show that the image file size of $40\times$ WSI images can be reduced by a factor of over 6000 with negligible loss of glomerulus detection accuracy.