检索结果-内蒙古大学图书馆

Combining Self-Supervised Learning and Adversarial Training Based Domain Adaptation for Speaker Verification

学校读者我要写书评

暂无评论

Combining Self-Supervised Learning and Adversarial Training ...

International Symposium on Chinese Spoken Language Processing

作者： Zhengyang Chen Shuai Wang Bing Han Yanmin Qian Department of Computer Science and Engineering Auditory Cognition and Computational Acoustics Lab MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai Shenzhen Research Institute of Big Data Shenzhen Chinese University of Hong Kong Shenzhen

ISBN: (数字)9798331516826

ISBN: (纸本)9798331516833

Adapting an existing well-trained system to a new domain using only unlabeled data is a highly sought-after yet challenging task for speaker verification in real-world scenarios. In this paper, we study two different domain adaptation methods, the adversarial domain adaptation (ADA) and the self-supervised learning-based domain adaptation (SSDA). To facilitate the deployment of unsupervised adaptation methods in applications, we conduct a detailed analysis of the characteristics of both the ADA and SSDA adaptation strategies. Our findings indicate that the SSDA strategy's performance is highly influenced by the amount of target domain data, whereas the ADA strategy is relatively insensitive to data quantity. Furthermore, augmenting target domain data enhances SSDA system performance but diminishes ADA performance. To further enhance system performance, we explore the complementarity between ADA and SSDA. Our results demonstrate that ADA and SSDA complement each other. When both strategies are applied jointly, the best system achieves over 20.0% relative Equal Error Rate (EER) improvement on the Cnceleb evaluation set and over 35.0% relative average EER improvement on the SRE16 Cantonese and Tagalog evaluation set under domain mismatched conditions.

关键词： Training Error analysis System performance Self-supervised learning Speech processing

Towards Context-Aware Domain Generalization: Understanding the Benefits and Limits of Marginal Transfer Learning

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Müller, Jens Kühmichel, Lars Rohbeck, Martin Radev, Stefan T. Köthe, Ullrich Computer Vision and Learning Lab Heidelberg University Germany Division of Computational Genomics and Systems Genetics DKFZ Germany Department of Cognitive Science Rensselaer Polytechnic Institute NY United States

In this work, we analyze the conditions under which information about the context of an input X can improve the predictions of deep learning models in new domains. Following work in marginal transfer learning in Domain Generalization (DG), we formalize the notion of context as a permutation-invariant representation of a set of data points that originate from the same domain as the input itself. We offer a theoretical analysis of the conditions under which this approach can, in principle, yield benefits, and formulate two necessary criteria that can be easily verified in practice. Additionally, we contribute insights into the kind of distribution shifts for which the marginal transfer learning approach promises robustness. Empirical analysis shows that our criteria are effective in discerning both favorable and unfavorable scenarios. Finally, we demonstrate that we can reliably detect scenarios where a model is tasked with unwarranted extrapolation in out-of-distribution (OOD) domains, identifying potential failure cases. Consequently, we showcase a method to select between the most predictive and the most robust model, circumventing the well-known trade-off between predictive performance and robustness. © 2023, CC BY.

关键词： Deep learning

Weisfeiler and Leman go machine learning: the story so far

学校读者我要写书评

暂无评论

The Journal of Machine Learning Research

The Journal of Machine Learning Research 2023年第1期24卷 15865-15923页

作者： Christopher Morris Yaron Lipman Haggai Maron Bastian Rieck Nils M. Kriege Martin Grohe Matthias Fey Karsten Borgwardt Department of Computer Science RWTH Aachen University Aachen Germany Meta AI Research Department of Computer Science and Applied Mathematics Weizmann Institute of Science Rehovot Israel NVIDIA Research Tel Aviv Israel AIDOS Lab Institute of AI for Health Helmholtz Zentrum München and Technical University of Munich Munich Germany Faculty of Computer Science and Research Network Data Science University of Vienna Vienna Austria Kumo.AI Mountain View CA Machine Learning & Computational Biology Lab Department of Biosystems Science and Engineering ETH Zürich Basel Switzerland and Swiss Institute of Bioinformatics Lausanne Switzerland

In recent years, algorithms and neural architectures based on the Weisfeiler-Leman algorithm, a well-known heuristic for the graph isomorphism problem, have emerged as a powerful tool for machine learning with graphs and relational data. Here, we give a comprehensive overview of the algorithm's use in a machine-learning setting, focusing on the supervised regime. We discuss the theoretical background, show how to use it for supervised graph and node representation learning, discuss recent extensions, and outline the algorithm's connection to (permutation-)equivariant neural architectures. Moreover, we give an overview of current applications and future directions to stimulate further research.

关键词： machine learning for graphs graph neural networks Weisfeiler-Leman algorithm expressivity equivariance

LiteVLoc: Map-Lite Visual Localization for Image Goal Navigation

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Jiao, Jianhao He, Jinhao Liu, Changkun Aegidius, Sebastian Hu, Xiangcheng Braud, Tristan Kanoulas, Dimitrios The Robot Perception and Learning Lab Intelligent Robotics Department of Computer Science University College London Gower Street LondonWC1E 6BT United Kingdom Nansha District Guangzhou China The Department of Computer Science and Engineering HKUST Hong Kong The Department of Electronic an Computer Engineering HKUST Hong Kong The AI Centre Department of Computer Science University College London Gower Street LondonWC1E 6BT United Kingdom Archimedes/Athena RC Greece

This paper presents LiteVLoc, a hierarchical visual localization framework that uses a lightweight topo-metric map to represent the environment. The method consists of three sequential modules that estimate camera poses in a coarse-to-fine manner. Unlike mainstream approaches relying on detailed 3D representations, LiteVLoc reduces storage overhead by leveraging learning-based feature matching and geometric solvers for metric pose estimation. A novel dataset for the map-free relocalization task is also introduced. Extensive experiments including localization and navigation in both simulated and real-world scenarios have validate the system's performance and demonstrated its precision and efficiency for large-scale deployment. Code and data will be made publicly available at https://***/LiteVLoc. Copyright © 2024, The Authors. All rights reserved.

关键词： Visualization

Better Hit the Nail on the Head than Beat around the Bush: Removing Protected Attributes with a Single Projection

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Haghighatkhah, Pantea Fokkens, Antske Sommerauer, Pia Speckmann, Bettina Verbeek, Kevin TU Eindhoven Department of Mathematics and Computer Science Netherlands Vrije Universiteit Amsterdam Computational Linguistics Text Mining Lab Netherlands

Bias elimination and recent probing studies attempt to remove specific information from embedding spaces. Here it is important to remove as much of the target information as possible, while preserving any other information present. INLP is a popular recent method which removes specific information through iterative nullspace projections. Multiple iterations, however, increase the risk that information other than the target is negatively affected. We introduce two methods that find a single targeted projection: Mean Projection (MP, more efficient) and Tukey Median Projection (TMP, with theoretical guarantees). Our comparison between MP and INLP shows that (1) one MP projection removes linear separability based on the target and (2) MP has less impact on the overall space. Further analysis shows that applying random projections after MP leads to the same overall effects on the embedding space as the multiple projections of INLP. Applying one targeted (MP) projection hence is methodologically cleaner than applying multiple (INLP) projections that introduce random effects. © 2022, CC BY.

关键词： Iterative methods

Enhancing Diversity in Multi-Objective Feature Selection

学校读者我要写书评

暂无评论

Enhancing Diversity in Multi-Objective Feature Selection

Congress on Evolutionary Computation

作者： Sevil Zanjani Miyandoab Shahryar Rahnamayan Azam Asilian Bidgoli Sevda Ebrahimi Masoud Makrehchi Nature-Inspired Computational Intelligence (NICI) Lab Department of Electrical Computer and Software Engineering Ontario Tech University Oshawa ON Canada Department of Engineering Brock University St. Catharines ON Canada Paculty of Science Wilfrid Laurier University Waterloo ON Canada

ISBN: (数字)9798350308365

ISBN: (纸本)9798350308372

Feature selection plays a pivotal role in the data preprocessing and model-building pipeline, significantly en-hancing model performance, interpretability, and resource efficiency across diverse domains. In population-based optimization methods, the generation of diverse individuals holds utmost importance for adequately exploring the problem landscape, particularly in highly multi-modal multi-objective optimization problems. Our study reveals that, in line with findings from sev-eral prior research papers, commonly employed crossover and mutation operations lack the capability to generate high-quality diverse individuals and tend to become confined to limited areas around various local optima. This paper introduces an augmen-tation to the diversity of the population in the well-established multi-objective scheme of the genetic algorithm, NSGA-II. This enhancement is achieved through two key components: the genuine initialization method and the substitution of the worst individuals with new randomly generated individuals as a re-initialization approach in each generation. The proposed multi-objective feature selection method undergoes testing on twelve real-world classification problems, with the number of features ranging from 2,400 to nearly 50,000. The results demonstrate that replacing the last front of the population with an equivalent number of new random individuals generated using the genuine initialization method and featuring a limited number of features substantially improves the population's quality and, consequently, enhances the performance of the multi-objective algorithm.

关键词： Sociology Pipelines Optimization methods Evolutionary computation Feature extraction Data models Distance measurement

Flexible Tactile Sensing Using PDMS/TiO2 Based Capacitor with MOSFET Structure

学校读者我要写书评

暂无评论

Flexible Tactile Sensing Using PDMS/TiO2 Based Capacitor wit...

IEEE SENSORS

作者： Kamalesh Tripathy Dhanranjan Kumar Karthick Thiyagarajan Mitradip Bhattacharjee Department of Electrical Engineeringand Computer Science i-Lab Indian Institute of Science Education and Research Bhopal Bhopal Madhya Pradesh India Smart Sensing and Robotics Laboratory (SensR Lab) Centre for Advanced Manufacturing Technology School of Engineering Design and Built Environment Western Sydney University Penrith NSW Australia

ISBN: (数字)9798350363517

ISBN: (纸本)9798350363524

In this paper, we present a flexible capacitive tactile sensor based on a polydimethylsiloxane (PDMS) elastomer, which is well-suited for a wide range of applications. The sensor incorporates an n-channel depletion-mode MOSFET to achieve an extended gate architecture. To enhance the sensing performance, TiO 2 nanoparticles are incorporated as nanofillers, resulting in an improved dielectric material ideal for highly sensitive capacitive pressure and tactile sensors. The 99PDMS-01TiO 2 nanocomposite (comprising 99 wt % PDMS and 1 wt % TiO 2 ) is utilized for promising flexible tactile sensor applications due to the synergistic effects between the PDMS matrix and TiO 2 nanoparticles, which enhance both the mechanical and sensing properties of the composite. The fabricated sensor exhibits a sensitivity of $\mathbf{30} \boldsymbol{\mu}\mathbf{A}/\mathbf{kPa}$ within the range of 0-10.5 kPa and a sensitivity of $\mathbf{9} \boldsymbol{\mu}\mathbf{A}/\mathbf{kPa}$ in the range of 10.5-30 kPa.

关键词： Nanoparticles MOSFET Sensitivity Nanocomposites Capacitors Tactile sensors Logic gates Robustness Sensors Titanium dioxide

Efficient Leverage Score Sampling for Tensor Train Decomposition

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Bharadwaj, Vivek Rakhshan, Beheshteh T. Malik, Osman Asif Rabusseau, Guillaume Electrical Engineering and Computer Science Department UC Berkeley United States Computational Research Division Lawrence Berkeley National Lab United States Mila DIRO Université de Montréal Canada CIFAR AI Canada

Tensor Train (TT) decomposition is widely used in the machine learning and quantum physics communities as a popular tool to efficiently compress high-dimensional tensor data. In this paper, we propose an efficient algorithm to accelerate computing the TT decomposition with the Alternating Least Squares (ALS) algorithm relying on exact leverage scores sampling. For this purpose, we propose a data structure that allows us to efficiently sample from the tensor with time complexity logarithmic in the tensor size. Our contribution specifically leverages the canonical form of the TT decomposition. By maintaining the canonical form through each iteration of ALS, we can efficiently compute (and sample from) the leverage scores, thus achieving significant speed-up in solving each sketched least-square problem. Experiments on synthetic and real data on dense and sparse tensors demonstrate that our method outperforms SVD-based and ALS-based algorithms. © 2024, CC BY-NC-SA.

关键词： Tensors

IMPROVING EQUIVARIANT NETWORKS WITH PROBABILISTIC SYMMETRY BREAKING

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Lawrence, Hannah Portilheiro, Vasco Zhang, Yan Kaba, Sékou-Oumar Department of Electrical Engineering and Computer Science MIT United States Gatsby Computational Neuroscience Unit UCL United Kingdom Samsung – SAIT AI Lab Montreal Canada Mila – Quebec Artficial Intelligence Institute Canada McGill University Canada

Equivariance encodes known symmetries into neural networks, often enhancing generalization. However, equivariant networks cannot break symmetries: the output of an equivariant network must, by definition, have at least the same self-symmetries as the input. This poses an important problem, both (1) for prediction tasks on domains where self-symmetries are common, and (2) for generative models, which must break symmetries in order to reconstruct from highly symmetric latent spaces. This fundamental limitation can be addressed by considering equivariant conditional distributions, instead of equivariant functions. We present novel theoretical results that establish necessary and sufficient conditions for representing such distributions. Concretely, this representation provides a practical framework for breaking symmetries in any equivariant network via randomized canonicalization. Our method, SymPE (Symmetry-breaking Positional Encodings), admits a simple interpretation in terms of positional encodings. This approach expands the representational power of equivariant networks while retaining the inductive bias of symmetry, which we justify through generalization bounds. Experimental results demonstrate that SymPE significantly improves performance of group-equivariant and graph neural networks across diffusion models for graphs, graph autoencoders, and lattice spin system modeling. © 2025, CC BY-NC-SA.

关键词： Graph neural networks