Searcharxiv⌕ Search

arXiv subjects

Tao Wang

Publications and source records attributed to Tao Wang.

At least 505 records · Page 28Linked to original sources
arXiv2023

Core localized alpha-channeling via low frequency Alfven mode generation in reversed shear scenarios

A novel channel for fuel ions heating in tokamak core plasma is proposed and analyzed using nonlinear gyrokinetic theory. The channel is achieved via spontaneous decay of reversed shear Alfvén eigenmode (RSAE) into low frequency Alfvén modes (LFAM), which then heat fuel ions via collisionless ion Landau damping. The conditions for RSAE spontaneous decay are investigated, and the saturation level and the consequent fuel ion heating rate are also derived. The channel is expected to be crucial for future reactors operating under reversed shear configurations, where fusion alpha particles are generated in the tokamak core where the magnetic shear is typically reversed, and there is a dense RSAE spectrum due to the small alpha particle characteristic dimensionless orbits.

physics.plasm-ph↗
arXiv2023

PCDAL: A Perturbation Consistency-Driven Active Learning Approach for Medical Image Segmentation and Classification

In recent years, deep learning has become a breakthrough technique in assisting medical image diagnosis. Supervised learning using convolutional neural networks (CNN) provides state-of-the-art performance and has served as a benchmark for various medical image segmentation and classification. However, supervised learning deeply relies on large-scale annotated data, which is expensive, time-consuming, and even impractical to acquire in medical imaging applications. Active Learning (AL) methods have been widely applied in natural image classification tasks to reduce annotation costs by selecting more valuable examples from the unlabeled data pool. However, their application in medical image segmentation tasks is limited, and there is currently no effective and universal AL-based method specifically designed for 3D medical image segmentation. To address this limitation, we propose an AL-based method that can be simultaneously applied to 2D medical image classification, segmentation, and 3D medical image segmentation tasks. We extensively validated our proposed active learning method on three publicly available and challenging medical image datasets, Kvasir Dataset, COVID-19 Infection Segmentation Dataset, and BraTS2019 Dataset. The experimental results demonstrate that our PCDAL can achieve significantly improved performance with fewer annotations in 2D classification and segmentation and 3D segmentation tasks. The codes of this study are available at https://github.com/ortonwang/PCDAL.

eess.IV↗
arXiv2023

Thickness and temperature dependent damping in La$_{0.67}$Sr$_{0.33}$MnO$_{3}$ epitaxial films

The damping of La0.67Sr0.33MnO3 (LSMO) epitaxial films as a function of thickness at different temperatures was studied. The competition between two scattering types (\r{ho}-like and σ-like) with entirely distinct thickness and temperature dependencies resulted in complicated damping behavior. The behavior of σ-like damping in LSMO films is consistent with the behavior in magnetic metal films. However, because \r{ho}-like damping is sensitive to the fine electron structure near the Fermi surface, the distortion of the oxygen octahedra controlled by the film thickness is an important factor in controlling the damping. Our study demonstrates that the complexity of damping in LSMO epitaxial films is a consequence of strong-correlation effects, which are characteristic of complex transition-metal oxides.

cond-mat.mtrl-sci↗
arXiv2023

Origin of the Coulomb Pseudopotential

We address the outstanding problem of electron pairing in the presence of strong Coulomb repulsion at small to moderate values of the Coulomb parameter, $r_s \lesssim 2$, and demonstrate that the pseudopotential framework is fundamentally biased and uncontrolled. Instead, one has to break the net result into two distinctively different effects: the Fermi liquid renormalization factor and the change in the effective low-energy coupling. The latter quantity is shown to behave non-monotonically with an extremum at $ r_s\approx 0.75$. Within the random-phase approximation, Coulomb interaction starts to enhance the effective pairing coupling at $r_s >2$, and the suppression of the critical temperature is entirely due to the renormalized Fermi liquid properties. Leading vertex corrections change this picture only quantitatively. Our results call for radical reconsideration of the widely accepted repulsive pseudopotential approach and show the need for precise microscopic treatment of Coulomb interactions in the problem of superconducting instability.

cond-mat.supr-con↗
arXiv2023

The nanoscale imaging of the bulk polycrystalline material with the effects of depth of field and field of view based on x-ray free electron laser

Microscale imaging of mesoscale bulk materials under dynamic compression is important for understanding their properties. In this work, we study the effects of the depth of field (DoF) and field of view (FoV) of the optical lens and extract the scattered light of the region to be imaged within the bulk polycrystalline material based on the objective Bragg coherent diffraction imaging. We describe how the DoF and FoV quantitatively limit the diffraction volume, where the DoF and FoV limit the scattering region parallel and perpendicular to the direction of the light source respectively. We demonstrate this scheme by simulating the separate imaging of a submicron-sized crack region within a few μm-sized Si bulk material, and obtain a high imaging quality. This scheme enables imaging of selected regions within bulk polycrystalline materials with the resolution up to the order of 10 nm.

physics.optics↗
arXiv2023

Boosting Fast and High-Quality Speech Synthesis with Linear Diffusion

Denoising Diffusion Probabilistic Models have shown extraordinary ability on various generative tasks. However, their slow inference speed renders them impractical in speech synthesis. This paper proposes a linear diffusion model (LinDiff) based on an ordinary differential equation to simultaneously reach fast inference and high sample quality. Firstly, we employ linear interpolation between the target and noise to design a diffusion sequence for training, while previously the diffusion path that links the noise and target is a curved segment. When decreasing the number of sampling steps (i.e., the number of line segments used to fit the path), the ease of fitting straight lines compared to curves allows us to generate higher quality samples from a random noise with fewer iterations. Secondly, to reduce computational complexity and achieve effective global modeling of noisy speech, LinDiff employs a patch-based processing approach that partitions the input signal into small patches. The patch-wise token leverages Transformer architecture for effective modeling of global information. Adversarial training is used to further improve the sample quality with decreased sampling steps. We test proposed method with speech synthesis conditioned on acoustic feature (Mel-spectrograms). Experimental results verify that our model can synthesize high-quality speech even with only one diffusion step. Both subjective and objective evaluations demonstrate that our model can synthesize speech of a quality comparable to that of autoregressive models with faster synthesis speed (3 diffusion steps).

cs.SD↗
arXiv2023

GigaST: A 10,000-hour Pseudo Speech Translation Corpus

This paper introduces GigaST, a large-scale pseudo speech translation (ST) corpus. We create the corpus by translating the text in GigaSpeech, an English ASR corpus, into German and Chinese. The training set is translated by a strong machine translation system and the test set is translated by human. ST models trained with an addition of our corpus obtain new state-of-the-art results on the MuST-C English-German benchmark test set. We provide a detailed description of the translation process and verify its quality. We make the translated text data public and hope to facilitate research in speech translation. Additionally, we also release the training scripts on NeurST to make it easy to replicate our systems. GigaST dataset is available at https://st-benchmark.github.io/resources/GigaST.

cs.CL↗
arXiv2023

Multi-View Imputation and Cross-Attention Network Based on Incomplete Longitudinal and Multimodal Data for Conversion Prediction of Mild Cognitive Impairment

Predicting whether subjects with mild cognitive impairment (MCI) will convert to Alzheimer's disease is a significant clinical challenge. Longitudinal variations and complementary information inherent in longitudinal and multimodal data are crucial for MCI conversion prediction, but persistent issue of missing data in these data may hinder their effective application. Additionally, conversion prediction should be achieved in the early stages of disease progression in clinical practice, specifically at baseline visit (BL). Therefore, longitudinal data should only be incorporated during training to capture disease progression information. To address these challenges, a multi-view imputation and cross-attention network (MCNet) was proposed to integrate data imputation and MCI conversion prediction in a unified framework. First, a multi-view imputation method combined with adversarial learning was presented to handle various missing data scenarios and reduce imputation errors. Second, two cross-attention blocks were introduced to exploit the potential associations in longitudinal and multimodal data. Finally, a multi-task learning model was established for data imputation, longitudinal classification, and conversion prediction tasks. When the model was appropriately trained, the disease progression information learned from longitudinal data can be leveraged by BL data to improve MCI conversion prediction at BL. MCNet was tested on two independent testing sets and single-modal BL data to verify its effectiveness and flexibility in MCI conversion prediction. Results showed that MCNet outperformed several competitive methods. Moreover, the interpretability of MCNet was demonstrated. Thus, our MCNet may be a valuable tool in longitudinal and multimodal data analysis for MCI conversion prediction. Codes are available at https://github.com/Meiyan88/MCNET.

eess.IV↗
arXiv2023

ADD 2023: the Second Audio Deepfake Detection Challenge

Audio deepfake detection is an emerging topic in the artificial intelligence community. The second Audio Deepfake Detection Challenge (ADD 2023) aims to spur researchers around the world to build new innovative technologies that can further accelerate and foster research on detecting and analyzing deepfake speech utterances. Different from previous challenges (e.g. ADD 2022), ADD 2023 focuses on surpassing the constraints of binary real/fake classification, and actually localizing the manipulated intervals in a partially fake speech as well as pinpointing the source responsible for generating any fake audio. Furthermore, ADD 2023 includes more rounds of evaluation for the fake audio game sub-challenge. The ADD 2023 challenge includes three subchallenges: audio fake game (FG), manipulation region location (RL) and deepfake algorithm recognition (AR). This paper describes the datasets, evaluation metrics, and protocols. Some findings are also reported in audio deepfake detection tasks.

cs.SD↗
arXiv2023

Improving speech translation by fusing speech and text

In speech translation, leveraging multimodal data to improve model performance and address limitations of individual modalities has shown significant effectiveness. In this paper, we harness the complementary strengths of speech and text, which are disparate modalities. We observe three levels of modality gap between them, denoted by Modal input representation, Modal semantic, and Modal hidden states. To tackle these gaps, we propose \textbf{F}use-\textbf{S}peech-\textbf{T}ext (\textbf{FST}), a cross-modal model which supports three distinct input modalities for translation: speech, text, and fused speech-text. We leverage multiple techniques for cross-modal alignment and conduct a comprehensive analysis to assess its impact on speech translation, machine translation, and fused speech-text translation. We evaluate FST on MuST-C, GigaST, and newstest benchmark. Experiments show that the proposed FST achieves an average 34.0 BLEU on MuST-C En$\rightarrow$De/Es/Fr (vs SOTA +1.1 BLEU). Further experiments demonstrate that FST does not degrade on MT task, as observed in prior works. Instead, it yields an average improvement of 3.2 BLEU over the pre-trained MT model.

cs.CL↗
arXiv2023

RFR-WWANet: Weighted Window Attention-Based Recovery Feature Resolution Network for Unsupervised Image Registration

The Swin transformer has recently attracted attention in medical image analysis due to its computational efficiency and long-range modeling capability. Owing to these properties, the Swin Transformer is suitable for establishing more distant relationships between corresponding voxels in different positions in complex abdominal image registration tasks. However, the registration models based on transformers combine multiple voxels into a single semantic token. This merging process limits the transformers to model and generate coarse-grained spatial information. To address this issue, we propose Recovery Feature Resolution Network (RFRNet), which allows the transformer to contribute fine-grained spatial information and rich semantic correspondences to higher resolution levels. Furthermore, shifted window partitioning operations are inflexible, indicating that they cannot perceive the semantic information over uncertain distances and automatically bridge the global connections between windows. Therefore, we present a Weighted Window Attention (WWA) to build global interactions between windows automatically. It is implemented after the regular and cyclic shift window partitioning operations within the Swin transformer block. The proposed unsupervised deformable image registration model, named RFR-WWANet, detects the long-range correlations, and facilitates meaningful semantic relevance of anatomical structures. Qualitative and quantitative results show that RFR-WWANet achieves significant improvements over the current state-of-the-art methods. Ablation experiments demonstrate the effectiveness of the RFRNet and WWA designs. Our code is available at \url{https://github.com/MingR-Ma/RFR-WWANet}.

cs.CV↗
arXiv2023

Distract Your Attention: Multi-head Cross Attention Network for Facial Expression Recognition

We present a novel facial expression recognition network, called Distract your Attention Network (DAN). Our method is based on two key observations. Firstly, multiple classes share inherently similar underlying facial appearance, and their differences could be subtle. Secondly, facial expressions exhibit themselves through multiple facial regions simultaneously, and the recognition requires a holistic approach by encoding high-order interactions among local features. To address these issues, we propose our DAN with three key components: Feature Clustering Network (FCN), Multi-head cross Attention Network (MAN), and Attention Fusion Network (AFN). The FCN extracts robust features by adopting a large-margin learning objective to maximize class separability. In addition, the MAN instantiates a number of attention heads to simultaneously attend to multiple facial areas and build attention maps on these regions. Further, the AFN distracts these attentions to multiple locations before fusing the attention maps to a comprehensive one. Extensive experiments on three public datasets (including AffectNet, RAF-DB, and SFEW 2.0) verified that the proposed method consistently achieves state-of-the-art facial expression recognition performance. Code will be made available at https://github.com/yaoing/DAN.

cs.CV↗
arXiv2023

A new emission mode of PSR B1859+07

Previous studies have identified two emission modes in PSR B1859+07: a normal mode that has three prominent components in the average profile, with the trailing one being the brightest, and an anomalous mode (i.e. the A mode) where emissions seem to be shifted to an earlier phase. Within the normal mode, further analysis has revealed the presence of two sub-modes, i.e. the cW mode and cB mode, where the central component can appear either weak or bright. As for the anomalous mode, a new bright component emerges in the advanced phase while the bright trailing component in the normal mode disappears. New observations of PSR B1859+07 by using the Five-hundred-meter Aperture Spherical radio Telescope (FAST) have revealed the existence of a previously unknown emission mode, dubbed as the Af mode. In this mode, all emission components seen in the normal and anomalous modes are detected. Notably, the mean polarization profiles of both the A and Af modes exhibit an orthogonal polarization angle jump in the bright leading component. The polarization angles for the central component in the original normal mode follow two distinct orthogonal polarization modes in the A and Af modes respectively. The polarization angles for the trailing component show almost the same but a small systematic shift in the A and Af modes, roughly following the values for the cW and cB modes. Those polarization features of this newly detected emission mode imply that the anomalous mode A of PSR B1859+07 is not a result of ``phase shift" or ``swooshes" of normal components, but simply a result of the varying intensities of different profile components. Additionally, subpulse drifting has been detected in the leading component of the Af mode.

astro-ph.HE↗
arXiv2023

EventFormer: AU Event Transformer for Facial Action Unit Event Detection

Facial action units (AUs) play an indispensable role in human emotion analysis. We observe that although AU-based high-level emotion analysis is urgently needed by real-world applications, frame-level AU results provided by previous works cannot be directly used for such analysis. Moreover, as AUs are dynamic processes, the utilization of global temporal information is important but has been gravely ignored in the literature. To this end, we propose EventFormer for AU event detection, which is the first work directly detecting AU events from a video sequence by viewing AU event detection as a multiple class-specific sets prediction problem. Extensive experiments conducted on a commonly used AU benchmark dataset, BP4D, show the superiority of EventFormer under suitable metrics.

cs.CV↗
arXiv2023

A round-trip multi-band quantum access network

The quantum network makes use of the quantum states to transmit data, which will revolutionize classical communication and allow for some breakthrough applications. The quantum key distribution (QKD) is one prominent application of quantum networks, and can protect the data transmission through quantum mechanics. In this work, we propose an expandable and cost-effective quantum access network, in which the round-trip structure makes quantum states travel in a circle to carry the information, and the multi-band technique is proposed to support multi-user access. Based on the round-trip multi-band quantum access network, we realize multi-user secure key sharing through the continuous-variable QKD (CV-QKD) protocol. Due to the encoding characteristics of CV-QKD, the quadrature components in different frequency bands can be used to transmit key information for different users. The feasibility of this scheme is confirmed by comprehensive noise analysis, and is verified by a proof-of-principle experiment. The results show that each user can achieve excess noise suppression and 600 bps level secure key generation under 30 km standard fiber transmission. Such networks have the ability of multi-user access theoretically and could be expanded by plugging in simple modules. Therefore, it paves the way for near-term large-scale quantum secure networks.

quant-ph↗
arXiv2023

A Soft Coordination Method of Heterogeneous Devices in Distribution System Voltage Control

With the continuous increase of photovoltaic (PV) penetration, the voltage control interactions between newly installed PV inverters and previously deployed on-load tap-changer (OLTC) transformers become ever more significant. To achieve coordinated voltage regulation, current methods often rely on a decision-making algorithm to fully take over the control of all devices, requiring OLTC to give up its existing tap switching logic and execute corresponding upgrades. Aiming at bridging this gap, a soft coordination framework is proposed in this paper. Specifically, the decision-making commands are only applied on inverters, and OLTC that retains its own operation rule will be indirectly controlled by the changed system voltage, which is a result of appropriately adjusting inverters' Var output. The proposed method achieves the soft coordination by establishing a modified actor-critic algorithm to train a proxy model of inverters. The well-trained proxy model can properly adjust inverters' Var output to "softly" coordinate OLTC's tap operations, which finally attains coordinated voltage regulation and line loss minimization. Simulation results verify the superiority of our proposed method over traditional ones in coordinating heterogeneous devices for voltage control.

eess.SY↗
arXiv2023

GRIG: Few-Shot Generative Residual Image Inpainting

Image inpainting is the task of filling in missing or masked region of an image with semantically meaningful contents. Recent methods have shown significant improvement in dealing with large-scale missing regions. However, these methods usually require large training datasets to achieve satisfactory results and there has been limited research into training these models on a small number of samples. To address this, we present a novel few-shot generative residual image inpainting method that produces high-quality inpainting results. The core idea is to propose an iterative residual reasoning method that incorporates Convolutional Neural Networks (CNNs) for feature extraction and Transformers for global reasoning within generative adversarial networks, along with image-level and patch-level discriminators. We also propose a novel forgery-patch adversarial training strategy to create faithful textures and detailed appearances. Extensive evaluations show that our method outperforms previous methods on the few-shot image inpainting task, both quantitatively and qualitatively.

cs.CV↗