PyTorchKR Papers

최신 논문

총 2,461편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2412.05840

LVP-CLIP:Revisiting CLIP for Continual Learning with Label Vector Pool

Continual learning aims to update a model so that it can sequentially learn new tasks without forgetting previously acquired knowledge. Recent continual learning approaches often leverage the vision-language model CLIP for its high-dimensional feature space and cross-modality feature matching. Traditional CLIP-based classification methods identify the most similar text label for a test image by comparing their embeddings. However, these methods are sensitive to the quality of text phrases and less effective for classes lacking meaningful text labels. In this work, we rethink CLIP-based continual learning and introduce the concept of Label Vector Pool (LVP). LVP replaces text labels with training images as similarity references, eliminating the need for ideal text descriptions. We present three variations of LVP and evaluate their performance on class and domain incremental learning tasks. Leveraging CLIP's high dimensional feature space, LVP learning algorithms are task-order invariant. The new knowledge does not modify the old knowledge, hence, there is minimum forgetting. Different tasks can be learned independently and in parallel with low computational and memory demands. Experimental results show that proposed LVP-based methods outperform the current state-of-the-art baseline by a significant margin of 40.7%.

arXiv:2412.06418공식 코드 · Papers with Code

Continual Learning for Segment Anything Model Adaptation

Although the current different types of SAM adaptation methods have achieved promising performance for various downstream tasks, such as prompt-based ones and adapter-based ones, most of them belong to the one-step adaptation paradigm. In real-world scenarios, we are generally confronted with the dynamic scenario where the data comes in a streaming manner. Driven by the practical need, in this paper, we first propose a novel Continual SAM adaptation (CoSAM) benchmark with 8 different task domains and carefully analyze the limitations of the existing SAM one-step adaptation methods in the continual segmentation scenario. Then we propose a novel simple-yet-effective Mixture of Domain Adapters (MoDA) algorithm which utilizes the Global Feature Tokens (GFT) and Global Assistant Tokens (GAT) modules to help the SAM encoder extract well-separated features for different task domains, and then provide the accurate task-specific information for continual learning. Extensive experiments demonstrate that our proposed MoDA obviously surpasses the existing classic continual learning methods, as well as prompt-based and adapter-based approaches for continual segmentation. Moreover, after sequential learning on the CoSAM benchmark with diverse data distributions, our MoDA maintains highly competitive results in the natural image domain, approaching the zero-shot performance of the original SAM, demonstrating its superior capability in knowledge preservation. Notably, the proposed MoDA can be seamlessly integrated into various one-step adaptation methods of SAM, which can consistently bring obvious performance gains. Code is available at \url{https://github.com/yangjl1215/CoSAM}

arXiv:2412.04748공식 코드 · Papers with Code

Decomposed Distribution Matching in Dataset Condensation

Dataset Condensation (DC) aims to reduce deep neural networks training efforts by synthesizing a small dataset such that it will be as effective as the original large dataset. Conventionally, DC relies on a costly bi-level optimization which prohibits its practicality. Recent research formulates DC as a distribution matching problem which circumvents the costly bi-level optimization. However, this efficiency sacrifices the DC performance. To investigate this performance degradation, we decomposed the dataset distribution into content and style. Our observations indicate two major shortcomings of: 1) style discrepancy between original and condensed data, and 2) limited intra-class diversity of condensed dataset. We present a simple yet effective method to match the style information between original and condensed data, employing statistical moments of feature maps as well-established style indicators. Moreover, we enhance the intra-class diversity by maximizing the Kullback-Leibler divergence within each synthetic class, i.e., content. We demonstrate the efficacy of our method through experiments on diverse datasets of varying size and resolution, achieving improvements of up to 4.1% on CIFAR10, 4.2% on CIFAR100, 4.3% on TinyImageNet, 2.0% on ImageNet-1K, 3.3% on ImageWoof, 2.5% on ImageNette, and 5.5% in continual learning accuracy.

arXiv:2412.04190

Directed Structural Adaptation to Overcome Statistical Conflicts and Enable Continual Learning

Adaptive networks today rely on overparameterized fixed topologies that cannot break through the statistical conflicts they encounter in the data they are exposed to, and are prone to "catastrophic forgetting" as the network attempts to reuse the existing structures to learn new task. We propose a structural adaptation method, DIRAD, that can complexify as needed and in a directed manner without being limited by statistical conflicts within a dataset. We then extend this method and present the PREVAL framework, designed to prevent "catastrophic forgetting" in continual learning by detection of new data and assigning encountered data to suitable models adapted to process them, without needing task labels anywhere in the workflow. We show the reliability of the DIRAD in growing a network with high performance and orders-of-magnitude simpler than fixed topology networks; and demonstrate the proof-of-concept operation of PREVAL, in which continual adaptation to new tasks is observed while being able to detect and discern previously-encountered tasks.

arXiv:2412.03081

A new Time-decay Radiomics Integrated Network (TRINet) for short-term breast cancer risk prediction

To facilitate early detection of breast cancer, there is a need to develop short-term risk prediction schemes that can prescribe personalized/individualized screening mammography regimens for women. In this study, we propose a new deep learning architecture called TRINet that implements time-decay attention to focus on recent mammographic screenings, as current models do not account for the relevance of newer images. We integrate radiomic features with an Attention-based Multiple Instance Learning (AMIL) framework to weigh and combine multiple views for better risk estimation. In addition, we introduce a continual learning approach with a new label assignment strategy based on bilateral asymmetry to make the model more adaptable to asymmetrical cancer indicators. Finally, we add a time-embedded additive hazard layer to perform dynamic, multi-year risk forecasting based on individualized screening intervals. We used two public datasets, namely 8,528 patients from the American EMBED dataset and 8,723 patients from the Swedish CSAW dataset in our experiments. Evaluation results on the EMBED test set show that our approach significantly outperforms state-of-the-art models, achieving AUC scores of 0.851, 0.811, 0.796, 0.793, and 0.789 across 1-, 2-, to 5-year intervals, respectively. Our results underscore the importance of integrating temporal attention, radiomic features, time embeddings, bilateral asymmetry, and continual learning strategies, providing a more adaptive and precise tool for short-term breast cancer risk prediction.

arXiv:2412.02865

Memory-efficient Continual Learning with Neural Collapse Contrastive

Contrastive learning has significantly improved representation quality, enhancing knowledge transfer across tasks in continual learning (CL). However, catastrophic forgetting remains a key challenge, as contrastive based methods primarily focus on "soft relationships" or "softness" between samples, which shift with changing data distributions and lead to representation overlap across tasks. Recently, the newly identified Neural Collapse phenomenon has shown promise in CL by focusing on "hard relationships" or "hardness" between samples and fixed prototypes. However, this approach overlooks "softness", crucial for capturing intra-class variability, and this rigid focus can also pull old class representations toward current ones, increasing forgetting. Building on these insights, we propose Focal Neural Collapse Contrastive (FNC^2), a novel representation learning loss that effectively balances both soft and hard relationships. Additionally, we introduce the Hardness-Softness Distillation (HSD) loss to progressively preserve the knowledge gained from these relationships across tasks. Our method outperforms state-of-the-art approaches, particularly in minimizing memory reliance. Remarkably, even without the use of memory, our approach rivals rehearsal-based methods, offering a compelling solution for data privacy concerns.

arXiv:2412.00776

Learning Mamba as a Continual Learner

Continual learning (CL) aims to efficiently learn and accumulate knowledge from a data stream with different distributions. By formulating CL as a sequence prediction task, meta-continual learning (MCL) enables to meta-learn an efficient continual learner based on the recent advanced sequence models, e.g., Transformers. Although attention-free models (e.g., Linear Transformers) can ideally match CL's essential objective and efficiency requirements, they usually perform not well in MCL. Considering that the attention-free Mamba achieves excellent performances matching Transformers' on general sequence modeling tasks, in this paper, we aim to answer a question -- Can attention-free Mamba perform well on MCL? By formulating Mamba with selective state space models (SSMs) for MCL tasks, we propose to meta-learn Mamba as a continual learner, referred to as MambaCL. By incorporating selectivity regularization, we can effectively train MambaCL. Through comprehensive experiments across various CL tasks, we also explore how Mamba and other models perform in different MCL scenarios. Our experiments and analyses highlight the promising performance and generalization capabilities of Mamba in MCL.

arXiv:2412.02627

Continual Learning of Personalized Generative Face Models with Experience Replay

We introduce a novel continual learning problem: how to sequentially update the weights of a personalized 2D and 3D generative face model as new batches of photos in different appearances, styles, poses, and lighting are captured regularly. We observe that naive sequential fine-tuning of the model leads to catastrophic forgetting of past representations of the individual's face. We then demonstrate that a simple random sampling-based experience replay method is effective at mitigating catastrophic forgetting when a relatively large number of images can be stored and replayed. However, for long-term deployment of these models with relatively smaller storage, this simple random sampling-based replay technique also forgets past representations. Thus, we introduce a novel experience replay algorithm that combines random sampling with StyleGAN's latent space to represent the buffer as an optimal convex hull. We observe that our proposed convex hull-based experience replay is more effective in preventing forgetting than a random sampling baseline and the lower bound.

arXiv:2412.01004

Adaptive Rank, Reduced Forgetting: Knowledge Retention in Continual Learning Vision-Language Models with Dynamic Rank-Selective LoRA

We investigate whether the pre-trained knowledge of vision-language models (VLMs), such as CLIP, can be retained or even enhanced during continual learning (CL) while absorbing knowledge from a data stream. Existing methods often rely on additional reference data, isolated components for distribution or domain predictions, leading to high training costs, increased inference complexity, and limited improvement potential for pre-trained models. To address these challenges, we first comprehensively analyze the effects of parameter update locations and ranks on downstream adaptation and knowledge retention. Based on these insights, we propose Dynamic Rank-Selective Low Rank Adaptation (LoRA), a universal and efficient CL approach that adaptively assigns ranks to LoRA modules based on their relevance to the current data. Unlike prior methods, our approach continually enhances the pre-trained VLM by retaining both the pre-trained knowledge and the knowledge acquired during CL. Our approach eliminates the need for explicit domain or distribution prediction and additional reference data, enabling seamless integration of new tasks while preserving pre-trained capabilities. It also maintains the original architecture and deployment pipeline of the pre-trained model without incurring any additional inference overhead. Extensive experiments and analyses demonstrate that our method outperforms state-of-the-art approaches in continually absorbing knowledge of downstream tasks while retaining pre-trained knowledge.

arXiv:2412.00860

Deep evolving semi-supervised anomaly detection

The aim of this paper is to formalise the task of continual semi-supervised anomaly detection (CSAD), with the aim of highlighting the importance of such a problem formulation which assumes as close to real-world conditions as possible. After an overview of the relevant definitions of continual semi-supervised learning, its components, anomaly detection extension, and the training protocols; the paper introduces a baseline model of a variational autoencoder (VAE) to work with semi-supervised data along with a continual learning method of deep generative replay with outlier rejection. The results show that such a use of extreme value theory (EVT) applied to anomaly detection can provide promising results even in comparison to an upper baseline of joint training. The results explore the effects of how much labelled and unlabelled data is present, of which class, and where it is located in the data stream. Outlier rejection shows promising initial results where it often surpasses a baseline method of Elastic Weight Consolidation (EWC). A baseline for CSAD is put forward along with the specific dataset setups used for reproducability and testability for other practitioners. Future research directions include other CSAD settings and further research into efficient continual hyperparameter tuning.

arXiv:2412.02509

FCL-ViT: Task-Aware Attention Tuning for Continual Learning

Continual Learning (CL) involves adapting the prior Deep Neural Network (DNN) knowledge to new tasks, without forgetting the old ones. However, modern CL techniques focus on provisioning memory capabilities to existing DNN models rather than designing new ones that are able to adapt according to the task at hand. This paper presents the novel Feedback Continual Learning Vision Transformer (FCL-ViT) that uses a feedback mechanism to generate real-time dynamic attention features tailored to the current task. The FCL-ViT operates in two Phases. In phase 1, the generic image features are produced and determine where the Transformer should attend on the current image. In phase 2, task-specific image features are generated that leverage dynamic attention. To this end, Tunable self-Attention Blocks (TABs) and Task Specific Blocks (TSBs) are introduced that operate in both phases and are responsible for tuning the TABs attention, respectively. The FCL-ViT surpasses state-of-the-art performance on Continual Learning compared to benchmark methods, while retaining a small number of trainable DNN parameters.

arXiv:2412.02270

Sustainable Self-evolution Adversarial Training

With the wide application of deep neural network models in various computer vision tasks, there has been a proliferation of adversarial example generation strategies aimed at deeply exploring model security. However, existing adversarial training defense models, which rely on single or limited types of attacks under a one-time learning process, struggle to adapt to the dynamic and evolving nature of attack methods. Therefore, to achieve defense performance improvements for models in long-term applications, we propose a novel Sustainable Self-Evolution Adversarial Training (SSEAT) framework. Specifically, we introduce a continual adversarial defense pipeline to realize learning from various kinds of adversarial examples across multiple stages. Additionally, to address the issue of model catastrophic forgetting caused by continual learning from ongoing novel attacks, we propose an adversarial data replay module to better select more diverse and key relearning data. Furthermore, we design a consistency regularization strategy to encourage current defense models to learn more from previously trained ones, guiding them to retain more past knowledge and maintain accuracy on clean samples. Extensive experiments have been conducted to verify the efficacy of the proposed SSEAT defense method, which demonstrates superior defense performance and classification accuracy compared to competitors.

arXiv:2412.02125

Optimizing Latent Goal by Learning from Trajectory Preference

A glowing body of work has emerged focusing on instruction-following policies for open-world agents, aiming to better align the agent's behavior with human intentions. However, the performance of these policies is highly susceptible to the initial prompt, which leads to extra efforts in selecting the best instructions. We propose a framework named Preference Goal Tuning (PGT). PGT allows an instruction following policy to interact with the environment to collect several trajectories, which will be categorized into positive and negative samples based on preference. Then we use preference learning to fine-tune the initial goal latent representation with the categorized trajectories while keeping the policy backbone frozen. The experiment result shows that with minimal data and training, PGT achieves an average relative improvement of 72.0% and 81.6% over 17 tasks in 2 different foundation policies respectively, and outperforms the best human-selected instructions. Moreover, PGT surpasses full fine-tuning in the out-of-distribution (OOD) task-execution environments by 13.4%, indicating that our approach retains strong generalization capabilities. Since our approach stores a single latent representation for each task independently, it can be viewed as an efficient method for continual learning, without the risk of catastrophic forgetting or task interference. In short, PGT enhances the performance of agents across nearly all tasks in the Minecraft Skillforge benchmark and demonstrates robustness to the execution environment.

arXiv:2411.18788

Investigating Plausibility of Biologically Inspired Bayesian Learning in ANNs

Catastrophic forgetting has been the leading issue in the domain of lifelong learning in artificial systems. Current artificial systems are reasonably good at learning domains they have seen before; however, as soon as they encounter something new, they either go through a significant performance deterioration or if you try to teach them the new distribution of data, they forget what they have learned before. Additionally, they are also prone to being overly confident when performing inference on seen as well as unseen data, causing significant reliability issues when lives are at stake. Therefore, it is extremely important to dig into this problem and formulate an approach that will be continually adaptable as well as reliable. If we move away from the engineering domain of such systems and look into biological systems, we can realize that these very systems are very efficient at computing the reliance as well as the uncertainty of accurate predictions that further help them refine the inference in a life-long setting. These systems are not perfect; however, they do give us a solid understanding of the reasoning under uncertainty which takes us to the domain of Bayesian reasoning. We incorporate this Bayesian inference with thresholding mechanism as to mimic more biologically inspired models, but only at spatial level. Further, we reproduce a recent study on Bayesian Inference with Spiking Neural Networks for Continual Learning to compare against it as a suitable biologically inspired Bayesian framework. Overall, we investigate the plausibility of biologically inspired Bayesian Learning in artificial systems on a vision dataset, MNIST, and show relative performance improvement under the conditions when the model is forced to predict VS when the model is not.

arXiv:2411.19903

Incremental Multi-Scene Modeling via Continual Neural Graphics Primitives

Neural radiance fields (NeRF) have revolutionized photorealistic rendering of novel views for 3D scenes. Despite their growing popularity and efficiency as 3D resources, NeRFs face scalability challenges due to the need for separate models per scene and the cumulative increase in training time for multiple scenes. The potential for incrementally encoding multiple 3D scenes into a single NeRF model remains largely unexplored. To address this, we introduce Continual-Neural Graphics Primitives (C-NGP), a novel continual learning framework that integrates multiple scenes incrementally into a single neural radiance field. Using a generative replay approach, C-NGP adapts to new scenes without requiring access to old data. We demonstrate that C-NGP can accommodate multiple scenes without increasing the parameter count, producing high-quality novel-view renderings on synthetic and real datasets. Notably, C-NGP models all $8$ scenes from the Real-LLFF dataset together, with only a $2.2\%$ drop in PSNR compared to vanilla NeRF, which models each scene independently. Further, C-NGP allows multiple style edits in the same network.

arXiv:2411.19154

DESIRE: Dynamic Knowledge Consolidation for Rehearsal-Free Continual Learning

Continual learning aims to equip models with the ability to retain previously learned knowledge like a human. Recent work incorporating Parameter-Efficient Fine-Tuning has revitalized the field by introducing lightweight extension modules. However, existing methods usually overlook the issue of information leakage caused by the fact that the experiment data have been used in pre-trained models. Once these duplicate data are removed in the pre-training phase, their performance can be severely affected. In this paper, we propose a new LoRA-based rehearsal-free method named DESIRE. Our method avoids imposing additional constraints during training to mitigate catastrophic forgetting, thereby maximizing the learning of new classes. To integrate knowledge from old and new tasks, we propose two efficient post-processing modules. On the one hand, we retain only two sets of LoRA parameters for merging and propose dynamic representation consolidation to calibrate the merged feature representation. On the other hand, we propose decision boundary refinement to address classifier bias when training solely on new class data. Extensive experiments demonstrate that our method achieves state-of-the-art performance on multiple datasets and strikes an effective balance between stability and plasticity. Our code will be publicly available.

arXiv:2411.18919공식 코드 · Papers with Code

Federated Continual Graph Learning

Managing evolving graph data presents substantial challenges in storage and privacy, and training graph neural networks (GNNs) on such data often leads to catastrophic forgetting, impairing performance on earlier tasks. Despite existing continual graph learning (CGL) methods mitigating this to some extent, they rely on centralized architectures and ignore the potential of distributed graph databases to leverage collective intelligence. To this end, we propose Federated Continual Graph Learning (FCGL) to adapt GNNs across multiple evolving graphs under storage and privacy constraints. Our empirical study highlights two core challenges: local graph forgetting (LGF), where clients lose prior knowledge when adapting to new tasks, and global expertise conflict (GEC), where the global GNN exhibits sub-optimal performance in both adapting to new tasks and retaining old ones, arising from inconsistent client expertise during server-side parameter aggregation. To address these, we introduce POWER, a framework that preserves experience nodes with maximum local-global coverage locally to mitigate LGF, and leverages pseudo-prototype reconstruction with trajectory-aware knowledge transfer to resolve GEC. Experiments on various graph datasets demonstrate POWER's superiority over federated adaptations of CGL baselines and vision-centric federated continual learning approaches.

arXiv:2411.17917공식 코드 · Papers with Code

DECODE: Domain-aware Continual Domain Expansion for Motion Prediction

Motion prediction is critical for autonomous vehicles to effectively navigate complex environments and accurately anticipate the behaviors of other traffic participants. As autonomous driving continues to evolve, the need to assimilate new and varied driving scenarios necessitates frequent model updates through retraining. To address these demands, we introduce DECODE, a novel continual learning framework that begins with a pre-trained generalized model and incrementally develops specialized models for distinct domains. Unlike existing continual learning approaches that attempt to develop a unified model capable of generalizing across diverse scenarios, DECODE uniquely balances specialization with generalization, dynamically adjusting to real-time demands. The proposed framework leverages a hypernetwork to generate model parameters, significantly reducing storage requirements, and incorporates a normalizing flow mechanism for real-time model selection based on likelihood estimation. Furthermore, DECODE merges outputs from the most relevant specialized and generalized models using deep Bayesian uncertainty estimation techniques. This integration ensures optimal performance in familiar conditions while maintaining robustness in unfamiliar scenarios. Extensive evaluations confirm the effectiveness of the framework, achieving a notably low forgetting rate of 0.044 and an average minADE of 0.584 m, significantly surpassing traditional learning strategies and demonstrating adaptability across a wide range of driving conditions.

arXiv:2411.18320

Continual Learning in Machine Speech Chain Using Gradient Episodic Memory

Continual learning for automatic speech recognition (ASR) systems poses a challenge, especially with the need to avoid catastrophic forgetting while maintaining performance on previously learned tasks. This paper introduces a novel approach leveraging the machine speech chain framework to enable continual learning in ASR using gradient episodic memory (GEM). By incorporating a text-to-speech (TTS) component within the machine speech chain, we support the replay mechanism essential for GEM, allowing the ASR model to learn new tasks sequentially without significant performance degradation on earlier tasks. Our experiments, conducted on the LJ Speech dataset, demonstrate that our method outperforms traditional fine-tuning and multitask learning approaches, achieving a substantial error rate reduction while maintaining high performance across varying noise conditions. We showed the potential of our semi-supervised machine speech chain approach for effective and efficient continual learning in speech recognition.

arXiv:2411.17475

COBRA: A Continual Learning Approach to Vision-Brain Understanding

Vision-Brain Understanding (VBU) aims to extract visual information perceived by humans from brain activity recorded through functional Magnetic Resonance Imaging (fMRI). Despite notable advancements in recent years, existing studies in VBU continue to face the challenge of catastrophic forgetting, where models lose knowledge from prior subjects as they adapt to new ones. Addressing continual learning in this field is, therefore, essential. This paper introduces a novel framework called Continual Learning for Vision-Brain (COBRA) to address continual learning in VBU. Our approach includes three novel modules: a Subject Commonality (SC) module, a Prompt-based Subject Specific (PSS) module, and a transformer-based module for fMRI, denoted as MRIFormer module. The SC module captures shared vision-brain patterns across subjects, preserving this knowledge as the model encounters new subjects, thereby reducing the impact of catastrophic forgetting. On the other hand, the PSS module learns unique vision-brain patterns specific to each subject. Finally, the MRIFormer module contains a transformer encoder and decoder that learns the fMRI features for VBU from common and specific patterns. In a continual learning setup, COBRA is trained in new PSS and MRIFormer modules for new subjects, leaving the modules of previous subjects unaffected. As a result, COBRA effectively addresses catastrophic forgetting and achieves state-of-the-art performance in both continual learning and vision-brain reconstruction tasks, surpassing previous methods.