Добірка наукової літератури з теми "Domain Shift Robustness"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся зі списками актуальних статей, книг, дисертацій, тез та інших наукових джерел на тему "Domain Shift Robustness".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Статті в журналах з теми "Domain Shift Robustness":
1
Goodarzi, Payman, Andreas Schütze, and Tizian Schneider. "Comparison of different ML methods concerning prediction quality, domain adaptation and robustness." tm - Technisches Messen 89, no. 4 (February 25, 2022): 224–39. http://dx.doi.org/10.1515/teme-2021-0129.
Анотація:
Abstract Nowadays machine learning methods and data-driven models have been used widely in different fields including computer vision, biomedicine, and condition monitoring. However, these models show performance degradation when meeting real-life situations. Domain or dataset shift or out-of-distribution (OOD) prediction is mentioned as the reason for this problem. Especially in industrial condition monitoring, it is not clear when we should be concerned about domain shift and which methods are more robust against this problem. In this paper prediction results are compared for a conventional machine learning workflow based on feature extraction, selection, and classification/regression (FESC/R) and deep neural networks on two publicly available industrial datasets. We show that it is possible to visualize the possible shift in domain using feature extraction and principal component analysis. Also, experimental competition shows that the cross-domain validated results of FESC/R are comparable to the reported state-of-the-art methods. Finally, we show that the results for simple randomly selected validation sets do not correctly represent the model performance in real-world applications.
2
Xu, Minghao, Jian Zhang, Bingbing Ni, Teng Li, Chengjie Wang, Qi Tian, and Wenjun Zhang. "Adversarial Domain Adaptation with Domain Mixup." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 04 (April 3, 2020): 6502–9. http://dx.doi.org/10.1609/aaai.v34i04.6123.
Анотація:
Recent works on domain adaptation reveal the effectiveness of adversarial learning on filling the discrepancy between source and target domains. However, two common limitations exist in current adversarial-learning-based methods. First, samples from two domains alone are not sufficient to ensure domain-invariance at most part of latent space. Second, the domain discriminator involved in these methods can only judge real or fake with the guidance of hard label, while it is more reasonable to use soft scores to evaluate the generated images or features, i.e., to fully utilize the inter-domain information. In this paper, we present adversarial domain adaptation with domain mixup (DM-ADA), which guarantees domain-invariance in a more continuous latent space and guides the domain discriminator in judging samples' difference relative to source and target domains. Domain mixup is jointly conducted on pixel and feature level to improve the robustness of models. Extensive experiments prove that the proposed approach can achieve superior performance on tasks with various degrees of domain shift and data complexity.
3
Fan, Mengbao, Binghua Cao, and Guiyun Tian. "Enhanced Measurement of Paper Basis Weight Using Phase Shift in Terahertz Time-Domain Spectroscopy." Journal of Sensors 2017 (2017): 1–14. http://dx.doi.org/10.1155/2017/3520967.
Анотація:
THz time-domain spectroscopy has evolved as a noncontact, safe, and efficient technique for paper characterization. Our previous work adopted peak amplitude and delay time as features to determine paper basis weight using terahertz time-domain spectroscopy. However, peak amplitude and delay time tend to suffer from noises, resulting in degradation of accuracy and robustness. This paper proposes a noise-robust phase-shift based method to enhance measurements of paper basis weight. Based on Fresnel Formulae, the physical relationship between phase shift and paper basis weight is formulated theoretically neglecting multiple reflections in the case of normal incidence. The established formulation indicates that phase shift correlates linearly with paper basis weight intrinsically. Subsequently, paper sheets were stacked to fabricate the samples with different basis weights, and experimental results verified the developed mathematical formulation. Moreover, a comparison was made between phase shift, peak amplitude, and delay time with respect to linearity, accuracy, and noise robustness. The results show that phase shift is superior to the others.
4
Murala, Kranthi Kumar, Dr M. Kamaraju, and Dr K. Ramanjaneyulu. "Digital Fingerprinting In Encrypted Domain." INTERNATIONAL JOURNAL OF COMPUTERS & TECHNOLOGY 12, no. 1 (December 15, 2013): 3138–46. http://dx.doi.org/10.24297/ijct.v12i1.3360.
Анотація:
Digital fingerprinting is a method for protecting multimedia content from illegal redistribution and identified the colluders.In copy protection, a content seller embeds a unique identity as a watermark into the content before it is sold to a buyer. When an illegal copy is found, the seller can identify illegal users by extracting the fingerprint. In this proposing an anonymous fingerprinting  based on a homomorphic additive encryption scheme, it present a construction of anti-collision codes created using BIBD(Balanced incomplete block design) codes technique and dither technique which makes use of LFSR (linear feedback shift register) are used for improving the high robustness and Security.
5
Li, Qingchuan, Jiangxing Zheng, Wenfeng Tan, Xingshu Wang, and Yingwei Zhao. "Traffic Sign Detection: Appropriate Data Augmentation Method from the Perspective of Frequency Domain." Mathematical Problems in Engineering 2022 (December 7, 2022): 1–11. http://dx.doi.org/10.1155/2022/9571513.
Анотація:
This study introduces a challenge faced by CNN in the task of traffic sign detection: how to achieve robustness to distributional shift. At present, all kinds of CNN models rely on strong data augmentation methods to enrich samples and achieve robustness, such as Mosaic and Mixup. In this study, we note that these methods do not have similar effects on combating noise. We explore the performance of augmentation strategies against disturbance in different frequency bands and provide understanding from the Fourier analysis perspective. This understanding can provide a guidance for selecting data augmentation strategies for different detection tasks and benchmark datasets.
6
Aryal, Jagannath, and Bipul Neupane. "Multi-Scale Feature Map Aggregation and Supervised Domain Adaptation of Fully Convolutional Networks for Urban Building Footprint Extraction." Remote Sensing 15, no. 2 (January 13, 2023): 488. http://dx.doi.org/10.3390/rs15020488.
Анотація:
Automated building footprint extraction requires the Deep Learning (DL)-based semantic segmentation of high-resolution Earth observation images. Fully convolutional networks (FCNs) such as U-Net and ResUNET are widely used for such segmentation. The evolving FCNs suffer from the inadequate use of multi-scale feature maps in their backbone of convolutional neural networks (CNNs). Furthermore, the DL methods are not robust in cross-domain settings due to domain-shift problems. Two scale-robust novel networks, namely MSA-UNET and MSA-ResUNET, are developed in this study by aggregating the multi-scale feature maps in U-Net and ResUNET with partial concepts of the feature pyramid network (FPN). Furthermore, supervised domain adaptation is investigated to minimise the effects of domain-shift between the two datasets. The datasets include the benchmark WHU Building dataset and a developed dataset with 5× fewer samples, 4× lower spatial resolution and complex high-rise buildings and skyscrapers. The newly developed networks are compared to six state-of-the-art FCNs using five metrics: pixel accuracy, adjusted accuracy, F1 score, intersection over union (IoU), and the Matthews Correlation Coefficient (MCC). The proposed networks outperform the FCNs in the majority of the accuracy measures in both datasets. Compared to the larger dataset, the network trained on the smaller one shows significantly higher robustness in terms of adjusted accuracy (by 18%), F1 score (by 31%), IoU (by 27%), and MCC (by 29%) during the cross-domain validation of MSA-UNET. MSA-ResUNET shows similar improvements, concluding that the proposed networks when trained using domain adaptation increase the robustness and minimise the domain-shift between the datasets of different complexity.
7
S. Garea, Alberto S., Dora B. Heras, and Francisco Argüello. "TCANet for Domain Adaptation of Hyperspectral Images." Remote Sensing 11, no. 19 (September 30, 2019): 2289. http://dx.doi.org/10.3390/rs11192289.
Анотація:
The use of Convolutional Neural Networks (CNNs) to solve Domain Adaptation (DA) image classification problems in the context of remote sensing has proven to provide good results but at high computational cost. To avoid this problem, a deep learning network for DA in remote sensing hyperspectral images called TCANet is proposed. As a standard CNN, TCANet consists of several stages built based on convolutional filters that operate on patches of the hyperspectral image. Unlike the former, the coefficients of the filter are obtained through Transfer Component Analysis (TCA). This approach has two advantages: firstly, TCANet does not require training based on backpropagation, since TCA is itself a learning method that obtains the filter coefficients directly from the input data. Second, DA is performed on the fly since TCA, in addition to performing dimensional reduction, obtains components that minimize the difference in distributions of data in the different domains corresponding to the source and target images. To build an operating scheme, TCANet includes an initial stage that exploits the spatial information by providing patches around each sample as input data to the network. An output stage performing feature extraction that introduces sufficient invariance and robustness in the final features is also included. Since TCA is sensitive to normalization, to reduce the difference between source and target domains, a previous unsupervised domain shift minimization algorithm consisting of applying conditional correlation alignment (CCA) is conditionally applied. The results of a classification scheme based on CCA and TCANet show that the DA technique proposed outperforms other more complex DA techniques.
8
Griffiths, Matthew P., André J. M. Pugin, and Dariush Motazedian. "Estimating local slope in the time-frequency domain: Velocity-independent seismic imaging in the near surface." GEOPHYSICS 85, no. 5 (July 28, 2020): U99—U107. http://dx.doi.org/10.1190/geo2019-0753.1.
Анотація:
Seismic reflection processing for multicomponent data is very time consuming. To automatically streamline and shorten this process, a new approach for estimating the local event slope (local static shift) in the time-frequency domain is proposed and tested. The seismic event slope is determined by comparing the local phase content of Stockwell transformed signals. This calculation allows for noninterfering arrivals to be aligned by iteratively correcting trace by trace. Alternatively, the calculation can be used in a velocity-independent imaging framework with the possibility of exporting the determined time and velocities for each common midpoint gather, which leads to a more robust moveout correction. Synthetic models are used to test the robustness of the calculation and compare it directly to an existing method of local slope estimation. Compared to dynamic time warping, our method is more robust to noise but less robust to large time shifts, which limits our method to shorter geophone spacing. We apply the calculation to near-surface shear-wave data and compare it directly to semblance/normal-moveout processing. Examples demonstrate that the calculation yields an accurate local slope estimate and can produce sections of better or equal quality to sections processed using the conventional approach with much less user time input. It also serves as a first example of velocity-independent processing applied to near-surface reflection data.
9
Yang, Fengxiang, Zhun Zhong, Hong Liu, Zheng Wang, Zhiming Luo, Shaozi Li, Nicu Sebe, and Shin'ichi Satoh. "Learning to Attack Real-World Models for Person Re-identification via Virtual-Guided Meta-Learning." Proceedings of the AAAI Conference on Artificial Intelligence 35, no. 4 (May 18, 2021): 3128–35. http://dx.doi.org/10.1609/aaai.v35i4.16422.
Анотація:
Recent advances in person re-identification (re-ID) have led to impressive retrieval accuracy. However, existing re-ID models are challenged by the adversarial examples crafted by adding quasi-imperceptible perturbations. Moreover, re-ID systems face the domain shift issue that training and testing domains are not consistent. In this study, we argue that learning powerful attackers with high universality that works well on unseen domains is an important step in promoting the robustness of re-ID systems. Therefore, we introduce a novel universal attack algorithm called ``MetaAttack'' for person re-ID. MetaAttack can mislead re-ID models on unseen domains by a universal adversarial perturbation. Specifically, to capture common patterns across different domains, we propose a meta-learning scheme to seek the universal perturbation via the gradient interaction between meta-train and meta-test formed by two datasets. We also take advantage of a virtual dataset (PersonX), instead of real ones, to conduct meta-test. This scheme not only enables us to learn with more comprehensive variation factors but also mitigates the negative effects caused by biased factors of real datasets. Experiments on three large-scale re-ID datasets demonstrate the effectiveness of our method in attacking re-ID models on unseen domains. Our final visualization results reveal some new properties of existing re-ID systems, which can guide us in designing a more robust re-ID model. Code and supplemental material are available at \url{https://github.com/FlyingRoastDuck/MetaAttack_AAAI21}.
10
Sun, Haidong, Cheng Liu, Hao Zhang, Yanming Cheng, and Yongyin Qu. "Research on a Self-Coupling PID Control Strategy for a ZVS Phase-Shift Full-Bridge Converter." Mathematical Problems in Engineering 2021 (March 8, 2021): 1–9. http://dx.doi.org/10.1155/2021/6670382.
Анотація:
As an important part of the high-frequency switching power supply, the control accuracy of the phase-shift full-bridge converter directly affects the efficiency of the switching power supply. To improve the stability and antidisturbance ability of phase-shift control systems, this article presents a dual closed-loop control system based on Self-Coupling PID (SC-PID) control and applies the SC-PID control strategy to the voltage control of the phase-shift full-bridge converter. To begin with, in response to the contradiction of traditional PID, SC-PID breaks the limitation of PID control by introducing a new control idea instead of weighted summation of each gain, which fundamentally solves the contradiction between overshoot and rapidity. Then, using the dimension attributes between gains to develop new tuning rules to solve the system load disturbance, output voltage deviation from the reference value, and other problems, the purpose is to ensure the stability of the output voltage and improve the control effect. At the same time, the stability of the whole control system is analyzed in the complex frequency domain. Finally, with the same main circuit and parameters, three types of controllers are built separately, and using MATLAB for simulation comparison, the simulation results show that the control system based on SC-PID has better steady-state accuracy, faster response, and better robustness, which proves the feasibility of the SC-PID control idea.
Дисертації з теми "Domain Shift Robustness":
1
Mahmood, Hassan. "Domain shift robustness in deep learning models." Thesis, Edith Cowan University, Research Online, Perth, Western Australia, 2023. https://ro.ecu.edu.au/theses/2621.
Анотація:
The advances in machine learning and artificial intelligence are drastically improving our capabilities of solving very easy to extremely complex tasks using computational models. Although these models perform very well on a given data distribution used for training, when presented a data drawn from a different distribution during inference, they tend to degrade in performance. Data bias, within-domain, out-of-domain deviation, and overfitting to the specific data are some of the main challenges for learning based models. These challenges are prevalent in different imaging datasets, when we are getting different image modalities from a variety of different imaging sensors each with different intensity distribution. The analysis of these imaging data becomes more challenging when data is multi-vendor and collected at multiple sites at different time points under different protocols. For example, medical data from different sites, natural images under different environmental conditions (e.g. day and night, sunny and cloudy), data through different imaging modalities are challenging to deal with learning based models. These variations in the data during inference time degrade the performance of the models. On the contrary, biological brains are much better at handling such unseen circumstances. By taking advantage of the current understanding of biological structures and their functionality, we can aim toward making improvements in the existing methods to make them relatively more robust against unseen variations.
In this thesis, we investigate the effect of different types of domain shifts on deep learning based methods. We choose to analyse the performance of deep learning based models for various computer vision tasks i.e. image registration, image classification and image segmentation. The aim is to thoroughly probe into limitations of deep learning based models and to investigate that how learning based models can be made robust against domain shifts. In this study, our focus is on the specific case, when learning based methods have no access to the possible domain shifts as in practice it is not possible to know all the possible variation in the data during training phase.
To address the issue of different intensity distributions (within-domain shifts) in medical image data under the image registration paradigm, we investigated the effects of introducing perceptual and structural-based losses, in comparison with mean square and cross correlation-based losses in the training of deep learning based registration models. Image registration is an important computer vision technique that can also be used to precisely monitor disease progression and to analyse large-scale datasets in a high-throughput manner. Deep learning based image registration methods are mainly inspired by optical flow-based backbone architectures with the addition of spatial transformer networks. The optical flow algorithm assumes a certain constraint on the pixel values in consecutive frames, we argue that this assumption violates in case of different intensity distributions which in turn affects the performance of registration methods. We addressed the specific case for non-rigid registration in brain MRI images. By adding perceptual and structural losses we observe that the models become more robust towards change in intensity.
We then explored the effects of local pixel contrast extracted through modelling a module of the human visual system for saliency region detection in dynamic natural scenes under different illumination conditions. Based on the clear effectiveness of adding such a bio-inspired approach to the existing methods on the natural imaging dataset, we proposed a novel bio-inspired layer (NeDev) in deep neural networks that can greatly enhance the robustness and tolerance against out-of-domain intensity distribution in the case of medical images as well natural image datasets. This layer transforms the input image into a common image space which is computed by local pixel variance. We benchmark the performance of our approach on different datasets to show the efficacy of the proposed layer. Finally, we provide an application tool for the community that can help them label, apply active learning, perform segmentation and registration tasks on medical imaging datasets with a set of trained models.
This study provides a thorough analysis of the effects of different types of domain shifts on deep learning based methods by investigating the performance against major computer vision tasks i.e. image registration, image classification and image segmentation. Findings of this research study through the combination of the different pathways have led to the conclusion of effectiveness of structural and local pixel deviation as a defence against within-domain and out-of-domain shifts.
Частини книг з теми "Domain Shift Robustness":
1
Full, Peter M., Fabian Isensee, Paul F. Jäger, and Klaus Maier-Hein. "Studying Robustness of Semantic Segmentation Under Domain Shift in Cardiac MRI." In Statistical Atlases and Computational Models of the Heart. M&Ms and EMIDEC Challenges, 238–49. Cham: Springer International Publishing, 2021. http://dx.doi.org/10.1007/978-3-030-68107-4_24.
2
Full, Peter M., Fabian Isensee, Paul F. Jäger, and Klaus Maier-Hein. "Abstract: Studying Robustness of Semantic Segmentation under Domain Shift in Cardiac MRI." In Bildverarbeitung für die Medizin 2021, 269. Wiesbaden: Springer Fachmedien Wiesbaden, 2021. http://dx.doi.org/10.1007/978-3-658-33198-6_64.
3
Silici, Laura, Jerry Knox, Andy Rowe, and Suppiramaniam Nanthikesan. "Evaluating Transformational Adaptation in Smallholder Farming: Insights from an Evidence Review." In Transformational Change for People and the Planet, 187–202. Cham: Springer International Publishing, 2022. http://dx.doi.org/10.1007/978-3-030-78853-7_13.
Анотація:
AbstractThe literature on smallholder farming and climate change adaptation (CCA) has predominantly investigated the barriers to and determinants of farmer uptake of adaptation interventions. Although useful, this evidence fails to highlight the changes or persistence of adaptation responses over time. Studies usually adopt a narrow focus on incremental actions that provide limited insights into transformative adaptation pathways and how fundamental shifts in policy can address the root causes of vulnerability across different sectors and dimensions. Drawing on an evidence synthesis commissioned by the International Fund for Agricultural Development’s Independent Office of Evaluation, this chapter outlines how lessons from CCA interventions can be transferred via three learning domains that are essential for transformational change: scaling-up (in its multiple forms), knowledge management, and the human-environment nexus. We discuss the implications of our findings on monitoring, evaluation, and learning, highlighting the challenges that evaluators may face in capturing (a) the persistence or durability of transformational pathways, (b) the complexity of “super-wicked” problems, and (c) the relevance of context-dependent dynamics, within a landscape setting. We also address the contribution of evidence reviews to contemporary debates around development policy linked to climate change and agriculture, and the implications and value of such reviews to provide independent scientific rigor and robustness to conventional programmatic evaluations.
4
Frank, David, Keyan Fang, and Patrick Fonti. "Dendrochronology: Fundamentals and Innovations." In Stable Isotopes in Tree Rings, 21–59. Cham: Springer International Publishing, 2022. http://dx.doi.org/10.1007/978-3-030-92698-4_2.
Анотація:
AbstractThis chapter overviews long-standing foundations, methods, and concepts of dendrochronology, yet also pays attention to a few related paradigm shifts driven by isotope measurements in tree-rings. The basics of annual ring formation are first reviewed, followed by structural descriptions of tree-rings at the macroscopic-to-microscopic scale including earlywoodandlatewoodin conifers (gymnosperms) and hardwoods (angiosperms), as well as wood anatomical features. Numerous examples of inter-disciplinary applications connected to various tree-ring parameters are provided. With the foundation of tree-rings established, this chapter then describes the process and necessity for crossdating—the process by which each and every ring is assigned to a specific year. Methods and terminology related to field sampling also briefly described. The long-standing paradigm of site selection criteria—well shown to maximize common signals in tree-ring width datasets—is challenged in a brief discussion of newer tree-ring isotope literature demonstrating that robust chronologies with high signal-to-noise ratios can be obtained at non-ecotonal locations. Opportunities for isotope measurements to enable crossdating in otherwise challenging contexts are likewise highlighted. The chapter reviews a conceptual framework to disaggregate tree-ring time-series, with special attention to detrending and standardization methods used to mitigate tree-age/size related noise common to many applications such as dendroclimatic reconstruction. Some of the drivers of long-term trends in tree-ring isotope data such as the increase in the atmospheric concentration of CO2, age/size/height trends, and climate variation are presented along with related debates/uncertainties evident in literature in order to establish priorities for future investigations. The development of tree-ring chronologies and related quality control metrics used to assess the common signal and the variance of tree-ring data are described, along with the limitations in correlation based statistics to determine the robustness of tree-ring datasets particularly in the low frequency domain. These statistical methods will gain relevance as tree-ring isotope datasets increasingly approach sample replications and dataset structures typical for tree-ring width measurements.
5
Gu, Irene Yu-Hua, and Vasile Gui. "Joint Space-Time-Range Mean ShiftBased Image and Video Segmentation." In Advances in Image and Video Segmentation, 113–39. IGI Global, 2006. http://dx.doi.org/10.4018/978-1-59140-753-9.ch006.
Анотація:
This chapter addresses image and video segmentation by using mean shift-based filtering and segmentation. Mean shift is an effective and elegant method to directly seek the local modes (or, local maxima) of the probability density function without the requirement of actually estimating it. Mean shift is proportional to the normalized density gradient estimate, and is pointing to the local stationary point (or, local mode) of the density estimate at which it converges. A mean shift filter can be related to a domain filter, a range filter or a bilateral filter depending on the variable setting in the kernel, and also has its own strength due to its flexibility and statistical basis. In this chapter a variety of mean shift filtering approaches are described for image/video segmentation and nonlinear edge-preserving image smoothing. A joint space-time-range domain mean shift-based video segmentation approach is presented. Segmentation of moving/static objects/background is obtained through inter-frame mode-matching in consecutive frames and motion vector mode estimation. Newly appearing objects/regions in the current frame due to new foreground objects or uncovered background regions are segmented by intra-frame mode estimation. Examples of image/video segmentation are included to demonstrate the effectiveness and robustness of these methods. Pseudo codes of the algorithms are also included.
6
Guicking, Axel, Peter Tandler, and Thomas Grasse. "Supporting Synchronous Collaboration with Heterogeneous Devices." In Interdisciplinary Perspectives on E-Collaboration, 12–30. IGI Global, 2010. http://dx.doi.org/10.4018/978-1-61520-676-6.ch002.
Анотація:
The increasing availability of mobile devices in today’s business contexts raises the demand to shift the focus of groupware framework design. Instead of solely focusing on functional requirements of specific application domains or device characteristics, nonfunctional requirements need to be taken into account as well. Flexibility concerning the integration of devices and tailorability of the framework according to different usage contexts is essential for addressing device heterogeneity. Besides flexibility, in order to support the development of real-world applications involving heterogeneous devices, robustness and scalability concerns have to be addressed explicitly by the framework. This article presents Agilo, a groupware framework for synchronous collaboration. The framework incorporates approaches addressing flexibility, robustness, and scalability issues. The combination of these concerns makes it suitable for development of collaborative applications involving up to hundreds of users. As an example application, a commercial electronic meeting system is presented by illustrating typical usage scenarios, explaining applicationspecific requirements and describing the system design.
Тези доповідей конференцій з теми "Domain Shift Robustness":
1
Zemčík, Tomáš. "Pedestrian Detector Domain Shift Robustness Evaluation, And Domain Shift Error Mitigation Proposal." In STUDENT EEICT 2021. Brno: Fakulta elektrotechniky a komunikacnich technologii VUT v Brne, 2021. http://dx.doi.org/10.13164/eeict.2021.181.
2
Sei Zhen Khong and Michael Cantoni. "Time-domain ν-gap robustness analysis for shift-invariant systems." In 2013 IEEE 52nd Annual Conference on Decision and Control (CDC). IEEE, 2013. http://dx.doi.org/10.1109/cdc.2013.6760462.
3
Wang, Haoqing, and Zhi-Hong Deng. "Cross-Domain Few-Shot Classification via Adversarial Task Augmentation." In Thirtieth International Joint Conference on Artificial Intelligence {IJCAI-21}. California: International Joint Conferences on Artificial Intelligence Organization, 2021. http://dx.doi.org/10.24963/ijcai.2021/149.
Анотація:
Few-shot classification aims to recognize unseen classes with few labeled samples from each class. Many meta-learning models for few-shot classification elaborately design various task-shared inductive bias (meta-knowledge) to solve such tasks, and achieve impressive performance. However, when there exists the domain shift between the training tasks and the test tasks, the obtained inductive bias fails to generalize across domains, which degrades the performance of the meta-learning models. In this work, we aim to improve the robustness of the inductive bias through task augmentation. Concretely, we consider the worst-case problem around the source task distribution, and propose the adversarial task augmentation method which can generate the inductive bias-adaptive 'challenging' tasks. Our method can be used as a simple plug-and-play module for various meta-learning models, and improve their cross-domain generalization capability. We conduct extensive experiments under the cross-domain setting, using nine few-shot classification datasets: mini-ImageNet, CUB, Cars, Places, Plantae, CropDiseases, EuroSAT, ISIC and ChestX. Experimental results show that our method can effectively improve the few-shot classification performance of the meta-learning models under domain shift, and outperforms the existing works. Our code is available at https://github.com/Haoqing-Wang/CDFSL-ATA.
4
Han, Zhongyi, Xian-Jin Gui, Chaoran Cui, and Yilong Yin. "Towards Accurate and Robust Domain Adaptation under Noisy Environments." In Twenty-Ninth International Joint Conference on Artificial Intelligence and Seventeenth Pacific Rim International Conference on Artificial Intelligence {IJCAI-PRICAI-20}. California: International Joint Conferences on Artificial Intelligence Organization, 2020. http://dx.doi.org/10.24963/ijcai.2020/314.
Анотація:
In non-stationary environments, learning machines usually confront the domain adaptation scenario where the data distribution does change over time. Previous domain adaptation works have achieved great success in theory and practice. However, they always lose robustness in noisy environments where the labels and features of examples from the source domain become corrupted. In this paper, we report our attempt towards achieving accurate noise-robust domain adaptation. We first give a theoretical analysis that reveals how harmful noises influence unsupervised domain adaptation. To eliminate the effect of label noise, we propose an offline curriculum learning for minimizing a newly-defined empirical source risk. To reduce the impact of feature noise, we propose a proxy distribution based margin discrepancy. We seamlessly transform our methods into an adversarial network that performs efficient joint optimization for them, successfully mitigating the negative influence from both data corruption and distribution shift. A series of empirical studies show that our algorithm remarkably outperforms state of the art, over 10% accuracy improvements in some domain adaptation tasks under noisy environments.
5
Gobbi, Massimiliano, Gianpiero Mastinu, Augusto D’Orazio, Massimo Caudano, and Giorgio Faustini. "On the Optimisation of a Double Cone Synchroniser for Improved Manual Transmission Shiftability." In ASME 2002 International Mechanical Engineering Congress and Exposition. ASMEDC, 2002. http://dx.doi.org/10.1115/imece2002-32911.
Анотація:
The paper presents a method to optimise the synchroniser of a road vehicle gearbox in order to improve shiftability and driver comfort. A multi-body physical model of the synchroniser has been developed and validated experimentally. The optimisation method is based on a Multi-objective Programming approach, and it allows to tune the thirty-two parameters of the synchroniser in order to achieve the desired dynamic behaviour of the system during a reference shift action, defined by seven performance indices. A Global Approximation procedure has been followed to solve numerically the optimisation problem. A special study has been performed and implemented in order to explore all of the feasible design solutions within the design variables domain. A global sensitivity method has been applied in order to analyse the relationships among the thirty-two design variables and the seven performance indices. Pareto-optimal design solutions have been computed in a very short time. These Pareto-optimal solutions have been checked for robustness by applying the minimum sensitivity method. The optimisation method has been applied with successful results. A number of optimised synchronisers have been defined, all of them featuring relevant improvements in the dynamic behaviour (shiftability) with respect to the reference synchroniser, aleady effective and under production.
6
Voulgaris, Georgios, Andrew Philippides, and Novi Quadrianto. "Deep Learning Robustness to Domain Shifts During Seasonal Variations." In IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium. IEEE, 2022. http://dx.doi.org/10.1109/igarss46834.2022.9883940.
7
Wijk, Hjalmar, Benjie Wang, and Marta Kwiatkowska. "Robustness Guarantees for Credal Bayesian Networks via Constraint Relaxation over Probabilistic Circuits." In Thirty-First International Joint Conference on Artificial Intelligence {IJCAI-22}. California: International Joint Conferences on Artificial Intelligence Organization, 2022. http://dx.doi.org/10.24963/ijcai.2022/677.
Анотація:
In many domains, worst-case guarantees on the performance (e.g. prediction accuracy) of a decision function subject to distributional shifts and uncertainty about the environment are crucial. In this work we develop a method to quantify the robustness of decision functions with respect to credal Bayesian networks, formal parametric models of the environment where uncertainty is expressed through credal sets on the parameters. In particular, we address the maximum marginal probability (MARmax) problem, that is, determining the greatest probability of an event (such as misclassification) obtainable for parameters in the credal set. We develop a method to faithfully transfer the problem into a constrained optimization problem on a probabilistic circuit. By performing a simple constraint relaxation, we show how to obtain a guaranteed upper bound on MARmax in linear time in the size of the circuit. We further theoretically characterize this constraint relaxation in terms of the original Bayesian network structure, which yields insight into the tightness of the bound. We implement the method and provide experimental evidence that the upper bound is often near tight and demonstrates improved scalability compared to other methods.