arXiv subjects

Xin Wang

Publications and source records attributed to Xin Wang.

At least 1,297 records · Page 72Linked to original sources

arXiv2022

Automatic speaker verification spoofing and deepfake detection using wav2vec 2.0 and data augmentation

The performance of spoofing countermeasure systems depends fundamentally upon the use of sufficiently representative training data. With this usually being limited, current solutions typically lack generalisation to attacks encountered in the wild. Strategies to improve reliability in the face of uncontrolled, unpredictable attacks are hence needed. We report in this paper our efforts to use self-supervised learning in the form of a wav2vec 2.0 front-end with fine tuning. Despite initial base representations being learned using only bona fide data and no spoofed data, we obtain the lowest equal error rates reported in the literature for both the ASVspoof 2021 Logical Access and Deepfake databases. When combined with data augmentation,these results correspond to an improvement of almost 90% relative to our baseline system.

eess.AS↗

arXiv2022

Synergistic Network Learning and Label Correction for Noise-robust Image Classification

Large training datasets almost always contain examples with inaccurate or incorrect labels. Deep Neural Networks (DNNs) tend to overfit training label noise, resulting in poorer model performance in practice. To address this problem, we propose a robust label correction framework combining the ideas of small loss selection and noise correction, which learns network parameters and reassigns ground truth labels iteratively. Taking the expertise of DNNs to learn meaningful patterns before fitting noise, our framework first trains two networks over the current dataset with small loss selection. Based on the classification loss and agreement loss of two networks, we can measure the confidence of training data. More and more confident samples are selected for label correction during the learning process. We demonstrate our method on both synthetic and real-world datasets with different noise types and rates, including CIFAR-10, CIFAR-100 and Clothing1M, where our method outperforms the baseline approaches.

cs.CV↗

arXiv2022

Data-driven Control of Dynamic Event-triggered Systems with Delays

This paper studies data-driven control of unknown sampled-data systems with communication delays under an event-triggering transmission mechanism. Data-based representations for time-invariant linear systems with known or unknown system input matrices are first developed, along with a novel class of dynamic triggering schemes for sampled-data systems with time delays. A model-based stability condition for the resulting event-triggered time-delay system is established using a looped-functional approach. Combining this model-based condition with the data-driven representations, data-based stability conditions are derived. Building on the data-based conditions, methods for co-designing the controller gain and the event-triggering matrix are subsequently provided for both cases with or without using the input matrix. Finally, numerical examples are presented to corroborate the usefulness of additional prior knowledge of the input matrix in reducing the conservatism of stability conditions, as well as the merits of the proposed data-driven event-triggered control schemes relative to existing results.

eess.SY↗

arXiv2022

Text-to-Speech Synthesis Techniques for MIDI-to-Audio Synthesis

Speech synthesis and music audio generation from symbolic input differ in many aspects but share some similarities. In this study, we investigate how text-to-speech synthesis techniques can be used for piano MIDI-to-audio synthesis tasks. Our investigation includes Tacotron and neural source-filter waveform models as the basic components, with which we build MIDI-to-audio synthesis systems in similar ways to TTS frameworks. We also include reference systems using conventional sound modeling techniques such as sample-based and physical-modeling-based methods. The subjective experimental results demonstrate that the investigated TTS components can be applied to piano MIDI-to-audio synthesis with minor modifications. The results also reveal the performance bottleneck -- while the waveform model can synthesize high quality piano sound given natural acoustic features, the conversion from MIDI to acoustic features is challenging. The full MIDI-to-audio synthesis system is still inferior to the sample-based or physical-modeling-based approaches, but we encourage TTS researchers to test their TTS models for this new task and improve the performance.

cs.SD↗

arXiv2022

An adaptive neuro-fuzzy model for attitude estimation and 2 control a 3 DOF system

In recent decades, one of the scientists' main concerns has been to improve the accuracy of satellite attitude, regardless of the expense. The obvious result is that a large number of control strategies have been used to address this problem. In this study, an adaptive neuro-fuzzy integrated (ANFIS) satellite attitude estimation and control system was developed. The controller is trained with the data provided by an optimal controller. A pulse modulator is used to generate the right ON/OFF commands of the thruster actuator. To evaluate the performance of the AN-FIS controller in closed-loop simulation, an ANFIS observer is used to estimate the attitude and angular velocities of the satellite using magnetometer, sun sensor and data gyro data. In addition, a new ANFIS system will be proposed and evaluated that can jointly control and estimate the system. The performance of the ANFIS controller is compared to the optimal PID controller in a Monte Carlo simulation with different initial conditions, disturbance and noise. The results show that the ANFIS controller can surpass the optimal PID controller in several aspects, including time and smoothness. In addition, the ANFIS estimator is examined and the results demonstrate the high ability of this designated observers. Both the control and estimation phases are simulated by a single ANFIS subsystem, taking into account the high capacity of ANFIS, and the results of using the ANFIS model are demonstrated.

eess.SY↗

arXiv2022

Near-optimal Top-k Pattern Mining

Nowadays, frequent pattern mining (FPM) on large graphs receives increasing attention, since it is crucial to a variety of applications, e.g., social analysis. Informally, the FPM problem is defined as finding all the patterns in a large graph with frequency above a user-defined threshold. However, this problem is nontrivial due to the unaffordable computational and space costs in the mining process. In light of this, we propose a cost-effective approach to mining near-optimal top-k patterns. Our approach applies a "level-wise" strategy to incrementally detect frequent patterns, hence is able to terminate as soon as top-k patterns are discovered. Moreover, we develop a technique to compute the lower bound of support with smart traverse strategy and compact data structures. Extensive experimental studies on real-life and synthetic graphs show that our approach performs well, i.e., it outperforms traditional counterparts in efficiency, memory footprint, recall and scalability.

cs.DB↗

arXiv2022

Lyman Continuum Galaxy Candidates in COSMOS

Star-forming galaxies are the sources likely to have reionized the universe. As we cannot observe them directly due to the opacity of the intergalactic medium at $z\gtrsim5$, we study $z\sim3\text{--}5$ galaxies as proxies to place observational constraints on cosmic reionization. Using new deep \textit{Hubble Space Telescope} rest-frame UV F336W and F435W imaging (30-orbit, $\sim40$~arcmin$^2$, $\sim29\text{--}30$~mag depth at 5$σ$), we attempt to identify a sample of Lyman continuum galaxies (LCGs). These are individual sources that emit ionizing flux below the Lyman break ($<912~\textÅ$). This population would allow us to constrain cosmic reionization parameters such as the number density and escape fraction ($f_{\rm esc}$) of ionizing sources. We compile a comprehensive parent sample that does not rely on the Lyman-break technique for redshifts. We present three new spectroscopic candidates at $z\sim3.7\text{--}4.4$, and 32 new photometric candidates. The high-resolution multi-band HST imaging and new Keck/Low Resolution Imaging Spectrometer (LRIS) redshifts make these promising spectroscopic LCG candidates. Using both a traditional and probabilistic approach, we find the most likely $f_{\rm esc}$ values for the three spectroscopic LCG candidates are $>100\%$, and therefore not physical. We are unable to confirm the true nature of these sources with the best available imaging and direct blue Keck/LRIS spectroscopy. More spectra, especially from the new class of 30 m telescopes, will be required to build a statistical sample of LCGs to place firm observational constraints on cosmic reionization.

astro-ph.GA↗

arXiv2022

Robust Attentive Deep Neural Network for Exposing GAN-generated Faces

GAN-based techniques that generate and synthesize realistic faces have caused severe social concerns and security problems. Existing methods for detecting GAN-generated faces can perform well on limited public datasets. However, images from existing public datasets do not represent real-world scenarios well enough in terms of view variations and data distributions (where real faces largely outnumber synthetic faces). The state-of-the-art methods do not generalize well in real-world problems and lack the interpretability of detection results. Performance of existing GAN-face detection models degrades significantly when facing imbalanced data distributions. To address these shortcomings, we propose a robust, attentive, end-to-end network that can spot GAN-generated faces by analyzing their eye inconsistencies. Specifically, our model learns to identify inconsistent eye components by localizing and comparing the iris artifacts between the two eyes automatically. Our deep network addresses the imbalance learning issues by considering the AUC loss and the traditional cross-entropy loss jointly. Comprehensive evaluations of the FFHQ dataset in terms of both balanced and imbalanced scenarios demonstrate the superiority of the proposed method.

cs.CV↗

arXiv2022

Fluorination Increases Hydrophobicity at the Macroscopic Level but not at the Microscopic Level

Hydrophobic interactions have been studied in detail in the past based on hydrophobic polymers, such as polystyrene (PS). Because fluorinated materials have relatively low surface energy, they often show both oleophobicity and hydrophobicity at the macroscopic level. However, it remains unknown how fluorination of hydrophobic polymer influences hydrophobicity at the microscopic level. In this work, we synthesized PS and fluorine-substituted PS (FPS) by reversible addition-fragmentation chain transfer polymerization method. Contact angle measurements confirmed that FPS is more hydrophobic than PS at the macroscopic level due to the introduction of fluorine. However, single molecule force spectroscopy experiments showed that the forces required to unfold the PS and FPS nanoparticles in water are indistinguishable, indicating that the strength of the hydrophobic ffect that drives the self-assembly of PS and FPS nanoparticles is the same at the microscopic level. The divergence of hydrophobic effect at the macroscopic and microscopic level may hint different underlying mechanisms: the hydrophobicity is dominated by the solvent hydration at the microscopic level and the surface-associated interaction at the macroscopic level.

cond-mat.soft↗

arXiv2022

Investigating self-supervised front ends for speech spoofing countermeasures

Self-supervised speech model is a rapid progressing research topic, and many pre-trained models have been released and used in various down stream tasks. For speech anti-spoofing, most countermeasures (CMs) use signal processing algorithms to extract acoustic features for classification. In this study, we use pre-trained self-supervised speech models as the front end of spoofing CMs. We investigated different back end architectures to be combined with the self-supervised front end, the effectiveness of fine-tuning the front end, and the performance of using different pre-trained self-supervised models. Our findings showed that, when a good pre-trained front end was fine-tuned with either a shallow or a deep neural network-based back end on the ASVspoof 2019 logical access (LA) training set, the resulting CM not only achieved a low EER score on the 2019 LA test set but also significantly outperformed the baseline on the ASVspoof 2015, 2021 LA, and 2021 deepfake test sets. A sub-band analysis further demonstrated that the CM mainly used the information in a specific frequency band to discriminate the bona fide and spoofed trials across the test sets.

eess.AS↗

arXiv2022

Estimating the confidence of speech spoofing countermeasure

Conventional speech spoofing countermeasures (CMs) are designed to make a binary decision on an input trial. However, a CM trained on a closed-set database is theoretically not guaranteed to perform well on unknown spoofing attacks. In some scenarios, an alternative strategy is to let the CM defer a decision when it is not confident. The question is then how to estimate a CM's confidence regarding an input trial. We investigated a few confidence estimators that can be easily plugged into a CM. On the ASVspoof2019 logical access database, the results demonstrate that an energy-based estimator and a neural-network-based one achieved acceptable performance in identifying unknown attacks in the test set. On a test set with additional unknown attacks and bona fide trials from other databases, the confidence estimators performed moderately well, and the CMs better discriminated bona fide and spoofed trials that had a high confidence score. Additional results also revealed the difficulty in enhancing a confidence estimator by adding unknown attacks to the training set.

eess.AS↗

arXiv2022

One dimensional reduced model for ITER relevant energetic particle transport

We set up a mapping procedure able to translate the evolution of the radial profile of fast ions, interacting with Toroidal Alfvén Eigenmodes, into the dynamics of an equivalent one dimensional bump-on-tail system. We apply this mapping technique to reproduce ITER relevant simulations, which clearly outlined deviations from the diffusive quasi-linear (QL) model. Our analysis demonstrates the capability of the one-dimensional beam-plasma dynamics to predict the relevant features of the non-linear hybrid LIGKA/HAGIS simulations. In particular, we clearly identify how the deviation from the QL evolutive profiles is due to the presence of avalanche processes. A detailed analysis regarding the reduced dimensionality is also addressed, by means of phase-space slicing based on constants of motion. In the conclusions, we outline the main criticalities and outcomes of the procedure, which must be satisfactorily addressed to make quantitative prediction on the observed outgoing fluxes in a Tokamak device.

physics.plasm-ph↗

arXiv2022

Accidental symmetries in the scalar potential of the Standard Model extended with two Higgs triplets

The extension of the Standard Model (SM) with two Higgs triplets offers an appealing way to account for both tiny Majorana neutrino masses via the type-II seesaw mechanism and the cosmological matter-antimatter asymmetry via the triplet leptogenesis. In this paper, we classify all possible accidental symmetries in the scalar potential of the two-Higgs-triplet model (2HTM). Based on the bilinear-field formalism, we show that the maximal symmetry group of the 2HTM potential is ${\rm SO(4)}$ and eight types of accidental symmetries in total can be identified. Furthermore, we examine the impact of the couplings between the SM Higgs doublet and the Higgs triplets on the accidental symmetries. The bounded-from-below conditions on the scalar potential with specific accidental symmetries are also derived. Taking the ${\rm SO(4)}$-invariant scalar potential as an example, we investigate the vacuum structures and the scalar mass spectra of the 2HTM.

hep-ph↗

arXiv2022

Decentralized Stochastic Proximal Gradient Descent with Variance Reduction over Time-varying Networks

In decentralized learning, a network of nodes cooperate to minimize an overall objective function that is usually the finite-sum of their local objectives, and incorporates a non-smooth regularization term for the better generalization ability. Decentralized stochastic proximal gradient (DSPG) method is commonly used to train this type of learning models, while the convergence rate is retarded by the variance of stochastic gradients. In this paper, we propose a novel algorithm, namely DPSVRG, to accelerate the decentralized training by leveraging the variance reduction technique. The basic idea is to introduce an estimator in each node, which tracks the local full gradient periodically, to correct the stochastic gradient at each iteration. By transforming our decentralized algorithm into a centralized inexact proximal gradient algorithm with variance reduction, and controlling the bounds of error sequences, we prove that DPSVRG converges at the rate of $O(1/T)$ for general convex objectives plus a non-smooth term with $T$ as the number of iterations, while DSPG converges at the rate $O(\frac{1}{\sqrt{T}})$. Our experiments on different applications, network topologies and learning models demonstrate that DPSVRG converges much faster than DSPG, and the loss function of DPSVRG decreases smoothly along with the training epochs.

cs.LG↗

arXiv2022

Dichotomic Pattern Mining with Applications to Intent Prediction from Semi-Structured Clickstream Datasets

We introduce a pattern mining framework that operates on semi-structured datasets and exploits the dichotomy between outcomes. Our approach takes advantage of constraint reasoning to find sequential patterns that occur frequently and exhibit desired properties. This allows the creation of novel pattern embeddings that are useful for knowledge extraction and predictive modeling. Finally, we present an application on customer intent prediction from digital clickstream data. Overall, we show that pattern embeddings play an integrator role between semi-structured data and machine learning models, improve the performance of the downstream task and retain interpretability.

cs.AI↗

arXiv2022

Adaptive Worker Grouping For Communication-Efficient and Straggler-Tolerant Distributed SGD

Wall-clock convergence time and communication load are key performance metrics for the distributed implementation of stochastic gradient descent (SGD) in parameter server settings. Communication-adaptive distributed Adam (CADA) has been recently proposed as a way to reduce communication load via the adaptive selection of workers. CADA is subject to performance degradation in terms of wall-clock convergence time in the presence of stragglers. This paper proposes a novel scheme named grouping-based CADA (G-CADA) that retains the advantages of CADA in reducing the communication load, while increasing the robustness to stragglers at the cost of additional storage at the workers. G-CADA partitions the workers into groups of workers that are assigned the same data shards. Groups are scheduled adaptively at each iteration, and the server only waits for the fastest worker in each selected group. We provide analysis and experimental results to elaborate the significant gains on the wall-clock time, as well as communication load and computation load, of G-CADA over other benchmark schemes.

cs.IT↗

arXiv2022

Robust Contrastive Learning against Noisy Views

Contrastive learning relies on an assumption that positive pairs contain related views, e.g., patches of an image or co-occurring multimodal signals of a video, that share certain underlying information about an instance. But what if this assumption is violated? The literature suggests that contrastive learning produces suboptimal representations in the presence of noisy views, e.g., false positive pairs with no apparent shared information. In this work, we propose a new contrastive loss function that is robust against noisy views. We provide rigorous theoretical justifications by showing connections to robust symmetric losses for noisy binary classification and by establishing a new contrastive bound for mutual information maximization based on the Wasserstein distance measure. The proposed loss is completely modality-agnostic and a simple drop-in replacement for the InfoNCE loss, which makes it easy to apply to existing contrastive frameworks. We show that our approach provides consistent improvements over the state-of-the-art on image, video, and graph contrastive learning benchmarks that exhibit a variety of real-world noise patterns.

cs.CV↗

arXiv2022

A Practical Guide to Logical Access Voice Presentation Attack Detection

Voice-based human-machine interfaces with an automatic speaker verification (ASV) component are commonly used in the market. However, the threat from presentation attacks is also growing since attackers can use recent speech synthesis technology to produce a natural-sounding voice of a victim. Presentation attack detection (PAD) for ASV, or speech anti-spoofing, is therefore indispensable. Research on voice PAD has seen significant progress since the early 2010s, including the advancement in PAD models, benchmark datasets, and evaluation campaigns. This chapter presents a practical guide to the field of voice PAD, with a focus on logical access attacks using text-to-speech and voice conversion algorithms and spoofing countermeasures based on artifact detection. It introduces the basic concept of voice PAD, explains the common techniques, and provides an experimental study using recent methods on a benchmark dataset. Code for the experiments is open-sourced.

eess.AS↗