Transfer_Learning
-
Classification of Brain Tumors using Hybrid Deep Learning Models
2025-08-02 12:56:18Neerav Nemchand GalaAbstract
The use of Convolutional Neural Networks (CNNs) has greatly improved the interpretation of medical images. However, conventional CNNs typically demand extensive computational resources and large training datasets. To address these limitations, this study applied transfer learning to achieve strong classification performance using fewer training samples. Specifically, the study compared EfficientNetV2 with its predecessor, EfficientNet, and with ResNet50 in classifying brain tumors into three types: glioma, meningioma, and pituitary tumors. Results showed that EfficientNetV2 delivered superior performance compared to the other models. However, this improvement came at the cost of increased training time, likely due to the model's greater complexity.
Abstract (translated)
卷积神经网络（CNN）的使用极大地提高了医学影像的解读能力。然而，传统的CNN通常需要大量的计算资源和大规模的训练数据集。为了克服这些限制，本研究应用了迁移学习，在较少的训练样本下实现了较强的分类性能。具体来说，该研究将EfficientNetV2与其前身EfficientNet以及ResNet50进行了比较，用于对大脑肿瘤进行三类（胶质瘤、脑膜瘤和垂体瘤）的分类。结果表明，EfficientNetV2相比其他模型表现出更优的性能。然而，这种改进是以训练时间增加为代价的，这可能是由于该模型复杂度更高所致。
URL
https://arxiv.org/abs/2508.01350
PDF
-
Enhancing material behavior discovery using embedding-oriented Physically-Guided Neural Networks with Internal Variables
2025-08-01 12:33:21Rub\'en Mu\~noz-Sierra, Manuel Doblar\'e, Jacobo Ayensa-Jim\'enezAbstract
Physically Guided Neural Networks with Internal Variables are SciML tools that use only observable data for training and and have the capacity to unravel internal state relations. They incorporate physical knowledge both by prescribing the model architecture and using loss regularization, thus endowing certain specific neurons with a physical meaning as internal state variables. Despite their potential, these models face challenges in scalability when applied to high-dimensional data such as fine-grid spatial fields or time-evolving systems. In this work, we propose some enhancements to the PGNNIV framework that address these scalability limitations through reduced-order modeling techniques. Specifically, we introduce alternatives to the original decoder structure using spectral decomposition, POD, and pretrained autoencoder-based mappings. These surrogate decoders offer varying trade-offs between computational efficiency, accuracy, noise tolerance, and generalization, while improving drastically the scalability. Additionally, we integrate model reuse via transfer learning and fine-tuning strategies to exploit previously acquired knowledge, supporting efficient adaptation to novel materials or configurations, and significantly reducing training time while maintaining or improving model performance. To illustrate these various techniques, we use a representative case governed by the nonlinear diffusion equation, using only observable data. Results demonstrate that the enhanced PGNNIV framework successfully identifies the underlying constitutive state equations while maintaining high predictive accuracy. It also improves robustness to noise, mitigates overfitting, and reduces computational demands. The proposed techniques can be tailored to various scenarios depending on data availability, resources, and specific modeling objectives, overcoming scalability challenges in all the scenarios.
Abstract (translated)
带有内部变量的物理引导神经网络（Physically Guided Neural Networks with Internal Variables，PGNNIV）是科学机器学习（SciML）工具的一种，它们仅使用可观察数据进行训练，并有能力揭示内部状态关系。这些模型通过规定模型架构和采用损失正则化来整合物理学知识，从而使某些特定的神经元具有作为内部状态变量的物理意义。尽管这类模型具备巨大潜力，但在应用于高维数据（如细网格空间场或时间演化系统）时面临可扩展性挑战。 在本工作中，我们提出了一些增强PGNNIV框架的方法以通过降阶建模技术解决这些可扩展性的限制。具体来说，我们引入了替代原始解码器结构的选项，包括光谱分解、POD（主成分分析）和基于预训练自动编码器映射的技术。这些替代解码器提供了计算效率、准确性、抗噪性以及泛化能力之间的不同权衡，并显著提高了可扩展性。 此外，我们通过迁移学习和微调策略整合了模型重用方法，利用先前获取的知识支持对新型材料或配置的高效适应，同时大幅减少训练时间并保持甚至提高模型性能。为了展示这些技术的应用效果，我们在一个由非线性扩散方程支配且仅使用可观察数据的情况下展示了代表性案例。 实验结果表明，增强后的PGNNIV框架能够成功识别潜在的状态方程式，并维持高预测准确性。它还提高了抗噪能力、缓解过拟合现象并减少计算需求。提出的这些技术可以根据数据可用性、资源和具体建模目标的不同情况进行调整，在所有情况下都能克服可扩展性的挑战。 通过这些方法，PGNNIV框架不仅能够处理更复杂的数据集和物理问题，而且能够在保持高精度的同时显著提高模型的效率和鲁棒性。这使得它成为科学研究中处理大规模数据和复杂系统的强大工具。
URL
https://arxiv.org/abs/2508.00959
PDF
-
Advancing Welding Defect Detection in Maritime Operations via Adapt-WeldNet and Defect Detection Interpretability Analysis
2025-08-01 07:19:23Kamal Basha S, Athira NambiarAbstract
Weld defect detection is crucial for ensuring the safety and reliability of piping systems in the oil and gas industry, especially in challenging marine and offshore environments. Traditional non-destructive testing (NDT) methods often fail to detect subtle or internal defects, leading to potential failures and costly downtime. Furthermore, existing neural network-based approaches for defect classification frequently rely on arbitrarily selected pretrained architectures and lack interpretability, raising safety concerns for deployment. To address these challenges, this paper introduces ``Adapt-WeldNet", an adaptive framework for welding defect detection that systematically evaluates various pre-trained architectures, transfer learning strategies, and adaptive optimizers to identify the best-performing model and hyperparameters, optimizing defect detection and providing actionable insights. Additionally, a novel Defect Detection Interpretability Analysis (DDIA) framework is proposed to enhance system transparency. DDIA employs Explainable AI (XAI) techniques, such as Grad-CAM and LIME, alongside domain-specific evaluations validated by certified ASNT NDE Level II professionals. Incorporating a Human-in-the-Loop (HITL) approach and aligning with the principles of Trustworthy AI, DDIA ensures the reliability, fairness, and accountability of the defect detection system, fostering confidence in automated decisions through expert validation. By improving both performance and interpretability, this work enhances trust, safety, and reliability in welding defect detection systems, supporting critical operations in offshore and marine environments.
Abstract (translated)
焊缝缺陷检测对于确保石油和天然气行业中管道系统的安全性和可靠性至关重要，尤其是在具有挑战性的海上和近海环境中。传统的非破坏性测试（NDT）方法往往无法检测到细微或内部缺陷，导致潜在的故障和高昂的停机成本。此外，现有的基于神经网络的缺陷分类方法通常依赖于任意选择的预训练架构，并缺乏可解释性，这引发了部署时的安全担忧。 为了解决这些问题，本文介绍了“Adapt-WeldNet”，这是一个用于焊接缺陷检测的自适应框架，该框架系统地评估了各种预训练架构、迁移学习策略和自适应优化器，以识别最佳性能模型及其超参数，并通过优化缺陷检测提供可操作见解。此外，还提出了一种新颖的缺陷检测可解释性分析（DDIA）框架，用于增强系统的透明度。DDIA采用可解释的人工智能（XAI）技术，如Grad-CAM和LIME，以及经过认证的ASNT NDE Level II专家验证的领域特定评估方法。通过结合“人机协作”（HITL）的方法，并遵循值得信赖的人工智能原则，DDIA确保了缺陷检测系统的可靠性、公平性和问责制，从而在专家验证的基础上增强对自动化决策的信心。 通过改进性能和可解释性，这项工作提高了焊接缺陷检测系统中的信任度、安全性和可靠性，支持海上和近海环境中关键操作的顺利进行。
URL
https://arxiv.org/abs/2508.00381
PDF
-
Boosting Vision Semantic Density with Anatomy Normality Modeling for Medical Vision-language Pre-training
2025-08-01 06:52:05Weiwei Cao, Jianpeng Zhang, Zhongyi Shui, Sinuo Wang, Zeli Chen, Xi Li, Le Lu, Xianghua Ye, Tingbo Liang, Qi Zhang, Ling ZhangAbstract
Vision-language pre-training (VLP) has great potential for developing multifunctional and general medical diagnostic capabilities. However, aligning medical images with a low signal-to-noise ratio (SNR) to reports with a high SNR presents a semantic density gap, leading to visual alignment bias. In this paper, we propose boosting vision semantic density to improve alignment effectiveness. On one hand, we enhance visual semantics through disease-level vision contrastive learning, which strengthens the model's ability to differentiate between normal and abnormal samples for each anatomical structure. On the other hand, we introduce an anatomical normality modeling method to model the distribution of normal samples for each anatomy, leveraging VQ-VAE for reconstructing normal vision embeddings in the latent space. This process amplifies abnormal signals by leveraging distribution shifts in abnormal samples, enhancing the model's perception and discrimination of abnormal attributes. The enhanced visual representation effectively captures the diagnostic-relevant semantics, facilitating more efficient and accurate alignment with the diagnostic report. We conduct extensive experiments on two chest CT datasets, CT-RATE and Rad-ChestCT, and an abdominal CT dataset, MedVL-CT69K, and comprehensively evaluate the diagnosis performance across multiple tasks in the chest and abdominal CT scenarios, achieving state-of-the-art zero-shot performance. Notably, our method achieved an average AUC of 84.9% across 54 diseases in 15 organs, significantly surpassing existing methods. Additionally, we demonstrate the superior transfer learning capabilities of our pre-trained model. Code is available at this https URL.
Abstract (translated)
视觉语言预训练（VLP）在开发多功能和通用的医学诊断能力方面具有巨大潜力。然而，将信号与噪声比（SNR）较低的医疗图像与SNR较高的报告对齐存在语义密度差距问题，导致视觉对齐偏差。本文中，我们提出了一种通过增强视觉语义密度来提高对齐效果的方法。一方面，我们通过疾病级别的视觉对比学习增强了视觉语义，这使模型能够更好地区分每个解剖结构的正常和异常样本的能力。另一方面，我们引入了解剖学正常性建模方法，为每一种解剖构造中的正常样本分布建模，并利用VQ-VAE在潜在空间中重构正常的视觉嵌入。这个过程通过利用异常样本中的分布变化来放大异常信号，从而增强了模型对异常属性的感知和识别能力。增强后的视觉表示有效地捕捉了与诊断相关的语义信息，有助于更高效、准确地将其与诊断报告进行对齐。 我们在两个胸部CT数据集（CT-RATE 和 Rad-ChestCT）以及一个腹部CT数据集（MedVL-CT69K）上进行了广泛的实验，并全面评估了多种任务中的诊断性能，在胸部和腹部CT场景中实现了最先进的零样本性能。值得注意的是，我们的方法在15个器官的54种疾病上的平均AUC达到84.9%，显著超过了现有方法的表现。此外，我们展示了预训练模型优越的迁移学习能力。 代码可在以下链接获取：[此 URL](https://this https URL)（请将“this https URL”替换为实际提供的链接地址）。
URL
https://arxiv.org/abs/2508.03742
PDF
-
Text-to-SQL Task-oriented Dialogue Ontology Construction
2025-07-31 09:08:59Renato Vukovic, Carel van Niekerk, Michael Heck, Benjamin Ruppik, Hsien-Chin Lin, Shutong Feng, Nurul Lubis, Milica GasicAbstract
Large language models (LLMs) are widely used as general-purpose knowledge sources, but they rely on parametric knowledge, limiting explainability and trustworthiness. In task-oriented dialogue (TOD) systems, this separation is explicit, using an external database structured by an explicit ontology to ensure explainability and controllability. However, building such ontologies requires manual labels or supervised training. We introduce TeQoDO: a Text-to-SQL task-oriented Dialogue Ontology construction method. Here, an LLM autonomously builds a TOD ontology from scratch without supervision using its inherent SQL programming capabilities combined with dialogue theory provided in the prompt. We show that TeQoDO outperforms transfer learning approaches, and its constructed ontology is competitive on a downstream dialogue state tracking task. Ablation studies demonstrate the key role of dialogue theory. TeQoDO also scales to allow construction of much larger ontologies, which we investigate on a Wikipedia and ArXiv dataset. We view this as a step towards broader application of ontologies to increase LLM explainability.
Abstract (translated)
大型语言模型（LLM）被广泛用作通用知识来源，但它们依赖于参数化知识，这限制了其可解释性和可信度。在任务导向对话（TOD）系统中，这种分离是明确的，使用一个由显式本体结构化的外部数据库来确保可解释性和可控性。然而，构建这样的本体需要手动标签或监督训练。 我们介绍了TeQoDO：一种利用LLM自身的SQL编程能力以及提示中的对话理论自主构建TOD本体的方法。这种方法不需要任何监督，并且从头开始创建本体。实验表明，TeQoDO在转移学习方法中表现更优，其构建的本体在下游对话状态跟踪任务上也具有竞争力。消融研究证明了对话理论的关键作用。此外，TeQoDO能够扩展以构建更大规模的本体，并且我们在维基百科和ArXiv数据集上对此进行了调查。 我们认为这是向更广泛地应用本体来增强LLM可解释性的一步。
URL
https://arxiv.org/abs/2507.23358
PDF
-
Whole-brain Transferable Representations from Large-Scale fMRI Data Improve Task-Evoked Brain Activity Decoding
2025-07-30 04:36:58Yueh-Po Peng, Vincent K. M. Cheung, Li SuAbstract
A fundamental challenge in neuroscience is to decode mental states from brain activity. While functional magnetic resonance imaging (fMRI) offers a non-invasive approach to capture brain-wide neural dynamics with high spatial precision, decoding from fMRI data -- particularly from task-evoked activity -- remains challenging due to its high dimensionality, low signal-to-noise ratio, and limited within-subject data. Here, we leverage recent advances in computer vision and propose STDA-SwiFT, a transformer-based model that learns transferable representations from large-scale fMRI datasets via spatial-temporal divided attention and self-supervised contrastive learning. Using pretrained voxel-wise representations from 995 subjects in the Human Connectome Project (HCP), we show that our model substantially improves downstream decoding performance of task-evoked activity across multiple sensory and cognitive domains, even with minimal data preprocessing. We demonstrate performance gains from larger receptor fields afforded by our memory-efficient attention mechanism, as well as the impact of functional relevance in pretraining data when fine-tuning on small samples. Our work showcases transfer learning as a viable approach to harness large-scale datasets to overcome challenges in decoding brain activity from fMRI data.
Abstract (translated)
神经科学的一个基本挑战是从脑活动解码心理状态。虽然功能性磁共振成像（fMRI）提供了一种非侵入性的方法，能够以高空间精度捕捉全脑的神经动态变化，但从fMRI数据中进行解码——特别是从任务诱发的活动中解码——仍然具有挑战性，原因包括其维度高、信噪比低以及受试者间的数据量有限。在这里，我们借鉴了计算机视觉领域的最新进展，并提出了STDA-SwiFT模型，这是一种基于变压器架构的模型，通过空间-时间分解注意力和自我监督对比学习的方法从大规模fMRI数据集中学习可迁移表示。 利用来自人类连接组项目（HCP）中995名受试者的预训练体素级表示，我们展示了我们的模型在解码任务诱发活动方面性能显著提升，尤其是在多个感觉和认知领域，并且即使进行最少的数据预处理亦如此。我们还表明了更大的感受野如何得益于我们记忆高效的注意力机制而带来的性能改进，以及当在小样本上微调时，预训练数据中的功能相关性的影响。 我们的研究展示了迁移学习作为一种可行的方法，可以利用大规模数据集来克服从fMRI数据解码脑活动的挑战。
URL
https://arxiv.org/abs/2507.22378
PDF
-
Prediction of acoustic field in 1-D uniform duct with varying mean flow and temperature using neural networks
2025-07-30 04:26:36D. Veerababu, Prasanta K. GhoshAbstract
Neural networks constrained by the physical laws emerged as an alternate numerical tool. In this paper, the governing equation that represents the propagation of sound inside a one-dimensional duct carrying a heterogeneous medium is derived. The problem is converted into an unconstrained optimization problem and solved using neural networks. Both the acoustic state variables: acoustic pressure and particle velocity are predicted and validated with the traditional Runge-Kutta solver. The effect of the temperature gradient on the acoustic field is studied. Utilization of machine learning techniques such as transfer learning and automatic differentiation for acoustic applications is demonstrated.
Abstract (translated)
由物理定律约束的神经网络作为一种替代数值工具应运而生。本文推导了描述一维管道中携带非均匀介质内声波传播的控制方程，并将其转化为无约束优化问题，利用神经网络进行求解。文中预测并验证了声学状态变量（即声压和质点速度），并与传统的龙格-库塔(Runge-Kutta)方法进行了对比。研究还探讨了温度梯度对声场的影响。此外，展示了在声学应用中使用机器学习技术如迁移学习和自动微分的利用情况。
URL
https://arxiv.org/abs/2507.22370
PDF
-
Measuring Time-Series Dataset Similarity using Wasserstein Distance
2025-07-29 19:33:10Hongjie Chen, Akshay Mehra, Josh Kimball, Ryan A. RossiAbstract
The emergence of time-series foundation model research elevates the growing need to measure the (dis)similarity of time-series datasets. A time-series dataset similarity measure aids research in multiple ways, including model selection, finetuning, and visualization. In this paper, we propose a distribution-based method to measure time-series dataset similarity by leveraging the Wasserstein distance. We consider a time-series dataset an empirical instantiation of an underlying multivariate normal distribution (MVN). The similarity between two time-series datasets is thus computed as the Wasserstein distance between their corresponding MVNs. Comprehensive experiments and visualization show the effectiveness of our approach. Specifically, we show how the Wasserstein distance helps identify similar time-series datasets and facilitates inference performance estimation of foundation models in both out-of-distribution and transfer learning evaluation, with high correlations between our proposed measure and the inference loss (>0.60).
Abstract (translated)
时间序列基础模型研究的出现提高了衡量时间序列数据集相似度的需求。一种时间序列数据集的相似性测量方法有助于多种研究方式，包括模型选择、微调和可视化等。在本文中，我们提出了一种基于分布的方法来利用Wasserstein距离（也称为地球移动距离）来衡量时间序列数据集之间的相似性。我们将一个时间序列数据集视为潜在多变量正态分布（MVN）的实证实例，并将两个时间序列数据集间的相似度计算为它们对应的MVNs之间的Wasser斯坦距离。 全面的实验和可视化展示了我们方法的有效性。具体而言，我们演示了Wasserstein距离如何帮助识别相似的时间序列数据集，并在分布外(OOD)评估及迁移学习中辅助基础模型推理性能估计，我们的衡量标准与推断损失之间具有高度的相关性（>0.60）。
URL
https://arxiv.org/abs/2507.22189
PDF
-
Boost Self-Supervised Dataset Distillation via Parameterization, Predefined Augmentation, and Approximation
2025-07-29 02:51:56Sheng-Feng Yu, Jia-Jiun Yao, Wei-Chen ChiuAbstract
Although larger datasets are crucial for training large deep models, the rapid growth of dataset size has brought a significant challenge in terms of considerable training costs, which even results in prohibitive computational expenses. Dataset Distillation becomes a popular technique recently to reduce the dataset size via learning a highly compact set of representative exemplars, where the model trained with these exemplars ideally should have comparable performance with respect to the one trained with the full dataset. While most of existing works upon dataset distillation focus on supervised datasets, we instead aim to distill images and their self-supervisedly trained representations into a distilled set. This procedure, named as Self-Supervised Dataset Distillation, effectively extracts rich information from real datasets, yielding the distilled sets with enhanced cross-architecture generalizability. Particularly, in order to preserve the key characteristics of original dataset more faithfully and compactly, several novel techniques are proposed: 1) we introduce an innovative parameterization upon images and representations via distinct low-dimensional bases, where the base selection for parameterization is experimentally shown to play a crucial role; 2) we tackle the instability induced by the randomness of data augmentation -- a key component in self-supervised learning but being underestimated in the prior work of self-supervised dataset distillation -- by utilizing predetermined augmentations; 3) we further leverage a lightweight network to model the connections among the representations of augmented views from the same image, leading to more compact pairs of distillation. Extensive experiments conducted on various datasets validate the superiority of our approach in terms of distillation efficiency, cross-architecture generalization, and transfer learning performance.
Abstract (translated)
尽管更大的数据集对于训练大型深度模型至关重要，但数据集规模的快速膨胀也带来了显著的成本挑战，这甚至导致了高昂的计算费用。为此，数据集蒸馏作为一种流行的技术被提出，旨在通过学习一组具有代表性的样本（称为示例）来减少数据集大小，在这种情况下，使用这些示例训练得到的模型应该与使用完整数据集训练出来的模型性能相当。然而，大多数现有的数据集蒸馏工作主要集中在监督学习的数据集上，我们则试图将图像及其自监督训练表示转化为一个浓缩后的集合。这一过程被称为自监督数据集蒸馏（Self-Supervised Dataset Distillation），它能够从真实世界的数据集中有效提取丰富信息，并生成具有增强跨架构泛化能力的浓缩集合。 为了更忠实且紧凑地保留原始数据集的关键特性，我们提出了几种新颖的技术： 1. 我们通过不同的低维基底引入了一种创新参数化的图像和表示方法，实验表明这种基础选择对参数化起着至关重要的作用； 2. 为了解决由数据增强的随机性所导致的不稳定问题——这是一个自监督学习中的关键组成部分，但在之前的自监督数据集蒸馏工作中被低估了——我们利用预定的数据增强策略来应对这一挑战； 3. 我们进一步通过一个轻量级网络建模来自同一图像的不同视图表示之间的连接关系，从而生成更为紧凑的蒸馏对。 在多个数据集上进行的广泛实验验证了我们在蒸馏效率、跨架构泛化能力以及迁移学习性能方面的优势。
URL
https://arxiv.org/abs/2507.21455
PDF
-
Evaluating Deep Learning Models for African Wildlife Image Classification: From DenseNet to Vision Transformers
2025-07-28 22:18:13Lukman Jibril Aliyu, Umar Sani Muhammad, Bilqisu Ismail, Nasiru Muhammad, Almustapha A Wakili, Seid Muhie Yimam, Shamsuddeen Hassan Muhammad, Mustapha AbdullahiAbstract
Wildlife populations in Africa face severe threats, with vertebrate numbers declining by over 65% in the past five decades. In response, image classification using deep learning has emerged as a promising tool for biodiversity monitoring and conservation. This paper presents a comparative study of deep learning models for automatically classifying African wildlife images, focusing on transfer learning with frozen feature extractors. Using a public dataset of four species: buffalo, elephant, rhinoceros, and zebra; we evaluate the performance of DenseNet-201, ResNet-152, EfficientNet-B4, and Vision Transformer ViT-H/14. DenseNet-201 achieved the best performance among convolutional networks (67% accuracy), while ViT-H/14 achieved the highest overall accuracy (99%), but with significantly higher computational cost, raising deployment concerns. Our experiments highlight the trade-offs between accuracy, resource requirements, and deployability. The best-performing CNN (DenseNet-201) was integrated into a Hugging Face Gradio Space for real-time field use, demonstrating the feasibility of deploying lightweight models in conservation settings. This work contributes to African-grounded AI research by offering practical insights into model selection, dataset preparation, and responsible deployment of deep learning tools for wildlife conservation.
Abstract (translated)
非洲野生动物种群正面临严重威胁，过去五十年间脊椎动物的数量下降了超过65%。为应对这一挑战，基于深度学习的图像分类技术已成为生物多样性监测和保护的一项有前景的工具。本文通过使用一个包含四种物种（水牛、大象、犀牛和斑马）的公开数据集，对比研究了几种深度学习模型在自动分类非洲野生动物图像方面的性能表现，重点关注冻结特征提取器的迁移学习方法。 实验中评估了DenseNet-201、ResNet-152、EfficientNet-B4以及Vision Transformer ViT-H/14四种模型的表现。其中，DenseNet-201在卷积神经网络（CNN）中的表现最佳，达到了67%的准确率；而ViT-H/14则实现了最高的总体准确率99%，但其计算成本显著更高，引发了部署方面的担忧。我们的实验突出了精度、资源需求与可部署性之间的权衡。 性能最好的卷积神经网络模型DenseNet-201被集成到了Hugging Face Gradio Space中，以支持实时现场应用，展示了在保护环境中部署轻量级模型的可行性。这项工作通过提供有关模型选择、数据集准备以及深度学习工具负责任部署的实际见解，为非洲本土人工智能研究做出了贡献。
URL
https://arxiv.org/abs/2507.21364
PDF
-
A Deep Learning Automatic Speech Recognition Model for Shona Language
2025-07-28 20:57:26Leslie Wellington Sirora, Mainford MutandavariAbstract
This study presented the development of a deep learning-based Automatic Speech Recognition system for Shona, a low-resource language characterized by unique tonal and grammatical complexities. The research aimed to address the challenges posed by limited training data, lack of labelled data, and the intricate tonal nuances present in Shona speech, with the objective of achieving significant improvements in recognition accuracy compared to traditional statistical models. The research first explored the feasibility of using deep learning to develop an accurate ASR system for Shona. Second, it investigated the specific challenges involved in designing and implementing deep learning architectures for Shona speech recognition and proposed strategies to mitigate these challenges. Lastly, it compared the performance of the deep learning-based model with existing statistical models in terms of accuracy. The developed ASR system utilized a hybrid architecture consisting of a Convolutional Neural Network for acoustic modelling and a Long Short-Term Memory network for language modelling. To overcome the scarcity of data, data augmentation techniques and transfer learning were employed. Attention mechanisms were also incorporated to accommodate the tonal nature of Shona speech. The resulting ASR system achieved impressive results, with a Word Error Rate of 29%, Phoneme Error Rate of 12%, and an overall accuracy of 74%. These metrics indicated the potential of deep learning to enhance ASR accuracy for under-resourced languages like Shona. This study contributed to the advancement of ASR technology for under-resourced languages like Shona, ultimately fostering improved accessibility and communication for Shona speakers worldwide.
Abstract (translated)
这项研究开发了一种基于深度学习的自动语音识别（ASR）系统，用于处理资源较少且具有独特音调和语法复杂性的绍纳语。该研究旨在通过使用深度学习克服训练数据有限、标签数据缺乏以及绍纳语特有的音调细微差别所带来的挑战，并实现与传统统计模型相比显著提高的识别准确性。具体来说，这项研究做了三件事： 1. 探索了利用深度学习为绍纳语开发准确ASR系统的可行性； 2. 研究了在设计和实施用于绍纳语音识别的深度学习架构时所面临的特定挑战，并提出了缓解这些挑战的战略方法； 3. 比较了基于深度学习模型与现有统计模型在准确性方面的表现。 该研究中开发的ASR系统采用了混合架构，其中包括用于声学建模的卷积神经网络（CNN）和用于语言建模的长短期记忆网络（LSTM）。为了解决数据稀缺的问题，研究人员采用了一系列的数据增强技术和迁移学习方法。此外，为了适应绍纳语特有的音调特征，还融入了注意力机制。 经过这些技术的应用，该ASR系统取得了显著成果：单词错误率为29%，音素错误率为12%，整体准确率达到了74%。这些指标表明深度学习在提高像绍纳这样的资源较少语言的自动语音识别准确性方面具有巨大潜力。 这项研究为推进用于绍纳及其他资源较少语言的自动语音识别技术做出了贡献，最终有助于提升全球绍纳语使用者之间的可访问性和交流效率。
URL
https://arxiv.org/abs/2507.21331
PDF
-
Multilingual Self-Taught Faithfulness Evaluators
2025-07-28 12:01:59Carlo Alfano, Aymen Al Marjani, Zeno Jonke, Amin Mantrach, Saab Mansour, Marcello FedericoAbstract
The growing use of large language models (LLMs) has increased the need for automatic evaluation systems, particularly to address the challenge of information hallucination. Although existing faithfulness evaluation approaches have shown promise, they are predominantly English-focused and often require expensive human-labeled training data for fine-tuning specialized models. As LLMs see increased adoption in multilingual contexts, there is a need for accurate faithfulness evaluators that can operate across languages without extensive labeled data. This paper presents Self-Taught Evaluators for Multilingual Faithfulness, a framework that learns exclusively from synthetic multilingual summarization data while leveraging cross-lingual transfer learning. Through experiments comparing language-specific and mixed-language fine-tuning approaches, we demonstrate a consistent relationship between an LLM's general language capabilities and its performance in language-specific evaluation tasks. Our framework shows improvements over existing baselines, including state-of-the-art English evaluators and machine translation-based approaches.
Abstract (translated)
大规模语言模型（LLM）的广泛应用增加了对自动评估系统的需求，特别是为了解决信息虚构的问题。尽管现有的忠实性评价方法显示出了一定的效果，但它们主要集中在英语上，并且通常需要昂贵的人工标注数据来微调专门化的模型。随着LLMs在多语种环境中得到更广泛的应用，迫切需要能够在无需大量标注数据的情况下跨语言准确评估忠实性的工具。本文提出了“Self-Taught Evaluators for Multilingual Faithfulness”框架，该框架仅从合成的多语言摘要数据中学习，并利用跨语言迁移学习技术。通过比较特定语言和混合语言微调方法的实验，我们展示了LLM的一般语言能力与其在特定语言评估任务中的表现之间存在一致的关系。我们的框架在现有的基线之上取得了改进，包括最先进的英语评价器和基于机器翻译的方法。
URL
https://arxiv.org/abs/2507.20752
PDF
-
Regularizing Subspace Redundancy of Low-Rank Adaptation
2025-07-28 11:52:56Yue Zhu, Haiwen Diao, Shang Gao, Jiazuo Yu, Jiawen Zhu, Yunzhi Zhuge, Shuai Hao, Xu Jia, Lu Zhang, Ying Zhang, Huchuan LuAbstract
Low-Rank Adaptation (LoRA) and its variants have delivered strong capability in Parameter-Efficient Transfer Learning (PETL) by minimizing trainable parameters and benefiting from reparameterization. However, their projection matrices remain unrestricted during training, causing high representation redundancy and diminishing the effectiveness of feature adaptation in the resulting subspaces. While existing methods mitigate this by manually adjusting the rank or implicitly applying channel-wise masks, they lack flexibility and generalize poorly across various datasets and architectures. Hence, we propose ReSoRA, a method that explicitly models redundancy between mapping subspaces and adaptively Regularizes Subspace redundancy of Low-Rank Adaptation. Specifically, it theoretically decomposes the low-rank submatrices into multiple equivalent subspaces and systematically applies de-redundancy constraints to the feature distributions across different projections. Extensive experiments validate that our proposed method consistently facilitates existing state-of-the-art PETL methods across various backbones and datasets in vision-language retrieval and standard visual classification benchmarks. Besides, as a training supervision, ReSoRA can be seamlessly integrated into existing approaches in a plug-and-play manner, with no additional inference costs. Code is publicly available at: this https URL.
Abstract (translated)
低秩适应（LoRA）及其变体在参数高效转移学习（PETL）中表现出强大的能力，通过减少可训练参数并利用重新参数化获益。然而，在训练过程中它们的投影矩阵仍然不受限制，导致表示冗余度高，并降低了所生成子空间中的特征适应效果。尽管现有的方法通过手动调整秩或隐式应用通道级掩码来缓解这一问题，但这些方法缺乏灵活性且在不同数据集和架构上泛化能力较差。 因此，我们提出了ReSoRA（Regularizing Subspace Redundancy of Low-Rank Adaptation），一种明确建模映射子空间之间冗余度并自适应地正则化低秩适应子空间冗余的方法。具体而言，它从理论上将低秩子矩阵分解为多个等价的子空间，并系统性地在不同投影下的特征分布上施加去冗余约束。 广泛的实验验证了我们提出的方法能够持续增强现有的最先进的PETL方法，在视觉-语言检索和标准视觉分类基准测试中的各种骨干网络和数据集上的表现。此外，作为训练监督，ReSoRA可以无缝集成到现有方法中，并且在不增加推理成本的情况下采用插件即用的方式。 代码可在以下网址获取：[this https URL]（请将URL替换为实际链接）。
URL
https://arxiv.org/abs/2507.20745
PDF
-
FAST: Similarity-based Knowledge Transfer for Efficient Policy Learning
2025-07-27 22:21:53Alessandro Capurso, Elia Piccoli, Davide BacciuAbstract
Transfer Learning (TL) offers the potential to accelerate learning by transferring knowledge across tasks. However, it faces critical challenges such as negative transfer, domain adaptation and inefficiency in selecting solid source policies. These issues often represent critical problems in evolving domains, i.e. game development, where scenarios transform and agents must adapt. The continuous release of new agents is costly and inefficient. In this work we challenge the key issues in TL to improve knowledge transfer, agents performance across tasks and reduce computational costs. The proposed methodology, called FAST - Framework for Adaptive Similarity-based Transfer, leverages visual frames and textual descriptions to create a latent representation of tasks dynamics, that is exploited to estimate similarity between environments. The similarity scores guides our method in choosing candidate policies from which transfer abilities to simplify learning of novel tasks. Experimental results, over multiple racing tracks, demonstrate that FAST achieves competitive final performance compared to learning-from-scratch methods while requiring significantly less training steps. These findings highlight the potential of embedding-driven task similarity estimations.
Abstract (translated)
迁移学习（TL）具有通过跨任务转移知识来加速学习的潜力。然而，它面临着诸如负向迁移、领域适应和在选择稳固源策略方面低效等关键挑战。这些问题往往代表了如游戏开发这类不断变化领域的关键问题，在这些领域中场景会转变且代理必须进行调整。新代理的持续推出既昂贵又效率低下。 在这项工作中，我们致力于解决TL中的核心问题以提高知识转移、跨任务代理性能，并减少计算成本。提出的名为FAST（基于自适应相似性的迁移框架）的方法利用视觉帧和文本描述来创建任务动态的潜在表示形式，进而用于估算环境间的相似性。这些相似度评分指导我们的方法选择候选策略从中学到迁移能力以简化新任务的学习过程。 实验结果显示，在多个赛车赛道上进行测试时，FAST实现了与从头开始学习方法相当甚至更好的最终性能，同时需要显著较少的训练步骤。这些发现突显了基于嵌入的任务相似性估算的潜力。
URL
https://arxiv.org/abs/2507.20433
PDF
-
Efficient Self-Supervised Neuro-Analytic Visual Servoing for Real-time Quadrotor Control
2025-07-26 09:17:38Sebastian Mocanu, Sebastian-Ion Nae, Mihai-Eugen Barbu, Marius LeordeanuAbstract
This work introduces a self-supervised neuro-analytical, cost efficient, model for visual-based quadrotor control in which a small 1.7M parameters student ConvNet learns automatically from an analytical teacher, an improved image-based visual servoing (IBVS) controller. Our IBVS system solves numerical instabilities by reducing the classical visual servoing equations and enabling efficient stable image feature detection. Through knowledge distillation, the student model achieves 11x faster inference compared to the teacher IBVS pipeline, while demonstrating similar control accuracy at a significantly lower computational and memory cost. Our vision-only self-supervised neuro-analytic control, enables quadrotor orientation and movement without requiring explicit geometric models or fiducial markers. The proposed methodology leverages simulation-to-reality transfer learning and is validated on a small drone platform in GPS-denied indoor environments. Our key contributions include: (1) an analytical IBVS teacher that solves numerical instabilities inherent in classical approaches, (2) a two-stage segmentation pipeline combining YOLOv11 with a U-Net-based mask splitter for robust anterior-posterior vehicle segmentation to correctly estimate the orientation of the target, and (3) an efficient knowledge distillation dual-path system, which transfers geometric visual servoing capabilities from the analytical IBVS teacher to a compact and small student neural network that outperforms the teacher, while being suitable for real-time onboard deployment.
Abstract (translated)
这项工作介绍了一种自监督神经分析、成本效益的模型，用于基于视觉的四旋翼飞行器控制。该模型中一个只有1.7M参数的小学生卷积网络（ConvNet）自动从改进后的基于图像的视觉伺服控制器（IBVS）这一分析教师那里学习。 我们的IBVS系统通过简化传统的视觉伺服方程来解决数值不稳定性，并实现了高效的稳定图像特征检测。通过知识蒸馏，学生的模型在推理速度上比老师的IBVS管道快11倍，同时表现出相似的控制精度，但在计算和内存成本方面显著降低。我们提出的仅依赖于视觉的自监督神经分析控制方法使四旋翼飞行器能够在不需要显式几何模型或标记点的情况下调整姿态与移动。 该方法利用了仿真到现实的迁移学习，并在一个小型无人机平台上通过GPS限制下的室内环境进行了验证。我们的主要贡献包括：(1)一种能够解决经典方法中固有数值不稳定的分析IBVS教师；(2)结合YOLOv11和基于U-Net的分割器的两阶段分割流水线，用于稳健地对前后车辆进行分割，从而准确估计目标的方向；(3)一个高效的双通道知识蒸馏系统，该系统能够将几何视觉伺服能力从分析IBVS教师转移到一个小而紧凑的学生神经网络上，该学生模型优于老师，并适用于实时机载部署。
URL
https://arxiv.org/abs/2507.19878
PDF
-
DOA: A Degeneracy Optimization Agent with Adaptive Pose Compensation Capability based on Deep Reinforcement Learning
2025-07-26 02:22:54Yanbin Li, Canran Xiao, Hongyang He, Shenghai Yuan, Zong Ke, Jiajie Yu, Zixiong Qin, Zhiguo Zhang, Wenzheng Chi, Wei ZhangAbstract
Particle filter-based 2D-SLAM is widely used in indoor localization tasks due to its efficiency. However, indoor environments such as long straight corridors can cause severe degeneracy problems in SLAM. In this paper, we use Proximal Policy Optimization (PPO) to train an adaptive degeneracy optimization agent (DOA) to address degeneracy problem. We propose a systematic methodology to address three critical challenges in traditional supervised learning frameworks: (1) data acquisition bottlenecks in degenerate dataset, (2) inherent quality deterioration of training samples, and (3) ambiguity in annotation protocol design. We design a specialized reward function to guide the agent in developing perception capabilities for degenerate environments. Using the output degeneracy factor as a reference weight, the agent can dynamically adjust the contribution of different sensors to pose optimization. Specifically, the observation distribution is shifted towards the motion model distribution, with the step size determined by a linear interpolation formula related to the degeneracy factor. In addition, we employ a transfer learning module to endow the agent with generalization capabilities across different environments and address the inefficiency of training in degenerate environments. Finally, we conduct ablation studies to demonstrate the rationality of our model design and the role of transfer learning. We also compare the proposed DOA with SOTA methods to prove its superior degeneracy detection and optimization capabilities across various environments.
Abstract (translated)
基于粒子滤波的2D-SLAM（同时定位与地图构建）方法由于其效率，在室内定位任务中被广泛应用。然而，如长直走廊等室内环境可能会导致SLAM产生严重的退化问题。本文中，我们采用近端策略优化（PPO）训练一个自适应退化解算器代理（DOA），以解决这种退化问题。文中提出了一种系统方法来应对传统监督学习框架中的三个关键挑战：1）退化数据集的数据获取瓶颈；2）训练样本固有的质量下降问题；3）注释协议设计的模糊性。 我们为该代理设计了一个专门化的奖励函数，以引导其在退化环境中发展感知能力。通过使用输出退化因子作为参考权重，代理能够动态调整不同传感器对姿态优化贡献的比例。具体而言，观察分布会向运动模型分布进行偏移，并且步长由与退化因子相关的线性插值公式确定。 此外，我们还采用了一个迁移学习模块，以赋予该代理跨不同环境的泛化能力，从而解决在退化环境中训练效率低下的问题。最后，通过消融研究证明了我们的模型设计的合理性以及迁移学习的作用，并且与最先进的方法进行了比较，验证了DOA在各种环境中的优越退化解算和优化能力。
URL
https://arxiv.org/abs/2507.19742
PDF
-
SAR-TEXT: A Large-Scale SAR Image-Text Dataset Built with SAR-Narrator and Progressive Transfer Learning
2025-07-24 18:45:30Xinjun Cheng, Yiguo He, Junjie Zhu, Chunping Qiu, Jun Wang, Qiangjuan Huang, Ke YangAbstract
Vision Language Models (VLMs) have achieved remarkable breakthroughs in the field of remote sensing in recent years. Synthetic Aperture Radar (SAR) imagery, with its all-weather capability, is essential in remote sensing, yet the lack of large-scale, high-quality SAR image-text datasets hinders its semantic understanding. In this paper, we construct SAR-Text, a large-scale and high-quality dataset consisting of over 130,000 SAR image-text pairs. To construct the SAR-Text dataset, we design the SAR-Narrator framework, which generates textual descriptions for SAR images through a multi-stage progressive transfer learning strategy. To verify the effectiveness of the SAR-TEXT dataset, we conduct experiments on three typical vision-language tasks: image-text retrieval, image captioning, and visual question answering (VQA). Specifically, we construct three representative models on SAR-TEXT: SAR-RS-CLIP, SAR-RS-CoCa, and SAR-GPT. SAR-RS-CLIP achieves notable improvements in retrieval performance, boosting average recall by 16.43% and 10.54% on the OSdataset-512 and HRSID test sets, respectively. In the captioning task, SAR-RS-CoCa achieves BLEU-4, SPICE, and CIDEr scores exceeding those of the original CoCa model by more than 8x, 4x, and 10x, respectively. In the VQA task, SAR-GPT outperforms baseline and single-stage models on multiple SAR-VQA datasets, demonstrating stronger semantic understanding and reasoning ability, as further confirmed by qualitative results. It is worth noting that, as a flexible captioning tool, SAR-Narrator can be readily adopted by the community to construct larger-scale SAR image-text datasets.
Abstract (translated)
近年来，视觉语言模型（VLMs）在遥感领域取得了显著的突破。合成孔径雷达（SAR）图像因其全天候能力，在遥感中至关重要，但缺乏大规模、高质量的SAR图像-文本数据集阻碍了其语义理解的进步。在这篇论文中，我们构建了一个名为SAR-Text的大规模高质量数据集，包含超过130,000对SAR图像和描述文本。为了构建SAR-Text数据集，我们设计了一种名为SAR-Narrator的框架，通过多阶段渐进式迁移学习策略为SAR图像生成文字描述。 为了验证SAR-TEXT数据集的有效性，我们在三个典型的视觉语言任务上进行了实验：图像-文本检索、图像字幕生成和视觉问答（VQA）。具体来说，在SAR-TEXT基础上构建了三种代表性模型：SAR-RS-CLIP、SAR-RS-CoCa和SAR-GPT。在图像-文本检索任务中，SAR-RS-CLIP实现了显著的性能提升，在OSdataset-512和HRSID测试集上分别提高了平均召回率16.43%和10.54%。在字幕生成任务中，SAR-RS-CoCa模型在BLEU-4、SPICE和CIDEr评分上均超过了原始CoCa模型8倍、4倍和10倍以上。而在视觉问答（VQA）任务中，SAR-GPT在多个SAR-VQA数据集上优于基准线模型和单阶段模型，展示了更强的语义理解和推理能力，这进一步通过定性结果得到了证实。 值得注意的是，作为一种灵活的文字生成工具，SAR-Narrator可以被社区轻松采用以构建更大规模的SAR图像-文本数据集。
URL
https://arxiv.org/abs/2507.18743
PDF
-
A Concept for Efficient Scalability of Automated Driving Allowing for Technical, Legal, Cultural, and Ethical Differences
2025-07-24 11:51:55Lars Ullrich, Michael Buchholz, Jonathan Petit, Klaus Dietmayer, Knut GraichenAbstract
Efficient scalability of automated driving (AD) is key to reducing costs, enhancing safety, conserving resources, and maximizing impact. However, research focuses on specific vehicles and context, while broad deployment requires scalability across various configurations and environments. Differences in vehicle types, sensors, actuators, but also traffic regulations, legal requirements, cultural dynamics, or even ethical paradigms demand high flexibility of data-driven developed capabilities. In this paper, we address the challenge of scalable adaptation of generic capabilities to desired systems and environments. Our concept follows a two-stage fine-tuning process. In the first stage, fine-tuning to the specific environment takes place through a country-specific reward model that serves as an interface between technological adaptations and socio-political requirements. In the second stage, vehicle-specific transfer learning facilitates system adaptation and governs the validation of design decisions. In sum, our concept offers a data-driven process that integrates both technological and socio-political aspects, enabling effective scalability across technical, legal, cultural, and ethical differences.
Abstract (translated)
自动驾驶（AD）的高效可扩展性是降低成本、增强安全、节约资源和最大化影响的关键。然而，目前的研究主要集中在特定车辆和情境上，而广泛的部署则需要在各种配置和环境中实现可扩展性。不同类型的车辆、传感器、执行器以及交通法规、法律要求、文化动态甚至伦理范式都对数据驱动开发的能力提出了高度的灵活性需求。本文旨在解决将通用能力向所需系统和环境进行可扩展适应性的挑战。我们的概念遵循一个两阶段微调过程。 在第一阶段，通过国家特定的奖励模型进行具体的环境微调，该模型充当技术适应与社会政治要求之间的接口。第二阶段则是车辆特异性的迁移学习，以促进系统的适应性，并管理设计决策的有效验证。 总而言之，我们提出的概念提供了一种数据驱动的过程，将技术和社会政治方面相结合，使跨技术、法律、文化和伦理差异的高效扩展成为可能。
URL
https://arxiv.org/abs/2507.18326
PDF
-
UrbanPulse: A Cross-City Deep Learning Framework for Ultra-Fine-Grained Population Transfer Prediction
2025-07-23 20:44:25Hongrong Yang, Markus SchlaepferAbstract
Accurate population flow prediction is essential for urban planning, transportation management, and public health. Yet existing methods face key limitations: traditional models rely on static spatial assumptions, deep learning models struggle with cross-city generalization, and Large Language Models (LLMs) incur high computational costs while failing to capture spatial structure. Moreover, many approaches sacrifice resolution by clustering Points of Interest (POIs) or restricting coverage to subregions, limiting their utility for city-wide analytics. We introduce UrbanPulse, a scalable deep learning framework that delivers ultra-fine-grained, city-wide OD flow predictions by treating each POI as an individual node. It combines a temporal graph convolutional encoder with a transformer-based decoder to model multi-scale spatiotemporal dependencies. To ensure robust generalization across urban contexts, UrbanPulse employs a three-stage transfer learning strategy: pretraining on large-scale urban graphs, cold-start adaptation, and reinforcement learning this http URL on over 103 million cleaned GPS records from three metropolitan areas in California, UrbanPulse achieves state-of-the-art accuracy and scalability. Through efficient transfer learning, UrbanPulse takes a key step toward making high-resolution, AI-powered urban forecasting deployable in practice across diverse cities.
Abstract (translated)
准确的人口流动预测对于城市规划、交通管理和公共卫生至关重要。然而，现有的方法面临一些关键限制：传统模型依赖于静态空间假设，深度学习模型在跨城市泛化方面遇到困难，而大型语言模型（LLMs）则因计算成本高昂且无法捕捉空间结构而受限。此外，许多方法通过聚类兴趣点（POIs）或仅限于次区域来牺牲分辨率，这限制了它们在整个城市的分析中的实用性。 我们引入了一种可扩展的深度学习框架UrbanPulse，它能够进行超精细的城市范围内的OD流量预测，并将每个POI视为一个独立节点。该框架结合了一个时间图卷积编码器和基于变压器的解码器来建模多尺度时空依赖性。为了确保在不同城市环境中的鲁棒泛化能力，UrbanPulse采用了一种三阶段的迁移学习策略：大规模城市图预训练、冷启动适应以及通过强化学习进行微调。 基于加州三个都会区超过1.03亿条经过清洗的GPS记录，UrbanPulse实现了最先进的准确性和可扩展性。通过高效的迁移学习，UrbanPulse为将高分辨率的人工智能驱动的城市预测在各种城市中实际部署迈出了关键一步。
URL
https://arxiv.org/abs/2507.17924
PDF
-
On the Interaction of Compressibility and Adversarial Robustness
2025-07-23 17:35:48Melih Barsbey, Ant\^onio H. Ribeiro, Umut \c{S}im\c{s}ekli, Tolga BirdalAbstract
Modern neural networks are expected to simultaneously satisfy a host of desirable properties: accurate fitting to training data, generalization to unseen inputs, parameter and computational efficiency, and robustness to adversarial perturbations. While compressibility and robustness have each been studied extensively, a unified understanding of their interaction still remains elusive. In this work, we develop a principled framework to analyze how different forms of compressibility - such as neuron-level sparsity and spectral compressibility - affect adversarial robustness. We show that these forms of compression can induce a small number of highly sensitive directions in the representation space, which adversaries can exploit to construct effective perturbations. Our analysis yields a simple yet instructive robustness bound, revealing how neuron and spectral compressibility impact $L_\infty$ and $L_2$ robustness via their effects on the learned representations. Crucially, the vulnerabilities we identify arise irrespective of how compression is achieved - whether via regularization, architectural bias, or implicit learning dynamics. Through empirical evaluations across synthetic and realistic tasks, we confirm our theoretical predictions, and further demonstrate that these vulnerabilities persist under adversarial training and transfer learning, and contribute to the emergence of universal adversarial perturbations. Our findings show a fundamental tension between structured compressibility and robustness, and suggest new pathways for designing models that are both efficient and secure.
Abstract (translated)
现代神经网络被期望同时满足一系列理想的特性，包括对训练数据的准确拟合、对未见输入的一般化能力、参数和计算效率以及对抗扰动的鲁棒性。尽管可压缩性和鲁棒性各自已经得到了广泛的研究，但它们相互作用的统一理解仍然难以捉摸。在本工作中，我们开发了一个原则性的框架来分析不同形式的可压缩性（如神经元级稀疏性和谱可压缩性）如何影响对抗鲁棒性。我们发现这些形式的压缩可以在表示空间中诱导出少量高度敏感的方向，这些方向可以被对手利用以构建有效的扰动。我们的分析得出了一种简单而有教育意义的鲁棒性界限，揭示了神经元和谱可压缩性是如何通过其对学习表示的影响来影响$L_\infty$ 和 $L_2$ 鲁棒性的。 尤为重要的是，我们所识别的脆弱性无论通过何种方式实现压缩都会出现——无论是通过正则化、架构偏置还是隐式学习动态。通过在合成和现实任务上的经验评估，我们确认了我们的理论预测，并进一步证明这些漏洞在对抗训练和迁移学习中仍然存在，而且有助于普遍对抗扰动的产生。 我们的发现表明，在结构化的可压缩性和鲁棒性之间存在着根本性的矛盾，并提出了设计既高效又安全模型的新路径。
URL
https://arxiv.org/abs/2507.17725
PDF
