Search

Previous studies have highlighted significant advancements in multimodal fusion. Nevertheless, such methods often encounter challenges regarding the efficacy of feature extraction, data integrity, consistency of feature dimensions, and adaptability across various downstream tasks. This paper proposes a generalized multimodal fusion method (GMF) via the Poisson-Nernst-Planck (PNP) equation, which adeptly addresses the aforementioned issues. Theoretically, the optimization objective for traditional multimodal tasks is formulated and redefined by integrating information entropy and the flow of gradient backward step. Leveraging these theoretical insights, the PNP equation is applied to feature fusion, rethinking multimodal features through the framework of charged particles in physics and controlling their movement through dissociation, concentration, and reconstruction. Building on these theoretical foundations, GMF disassociated features which extracted by the unimodal feature extractor into modality-specific and modality-invariant subspaces, thereby reducing mutual information and subsequently lowering the entropy of downstream tasks. The identifiability of the feature's origin enables our approach to function independently as a frontend, seamlessly integrated with a simple concatenation backend, or serve as a prerequisite for other modules. Experimental results on multiple downstream tasks show that the proposed GMF achieves performance close to the state-of-the-art (SOTA) accuracy while utilizing fewer parameters and computational resources. Furthermore, by integrating GMF with advanced fusion methods, we surpass the SOTA results., Comment: NeurIPS 2024 Rejected paper, 28 pages

Due to the lack of effective mpox detection tools, the mpox virus continues to spread worldwide and has once again been declared a public health emergency of international concern by the World Health Organization. Lightweight deep learning model-based detection systems are crucial to alleviate mpox outbreaks since they are suitable for widespread deployment, especially in resource-limited scenarios. However, the key to its successful application depends on ensuring that the model can effectively model local features and long-range dependencies in mpox lesions while maintaining lightweight. Inspired by the success of Mamba in modeling long-range dependencies and its linear complexity, we proposed a lightweight hybrid architecture called MpoxMamba for efficient mpox detection. MpoxMamba utilizes depth-wise separable convolutions to extract local feature representations in mpox skin lesions and greatly enhances the model's ability to model the global contextual information by grouped Mamba modules. Notably, MpoxMamba's parameter size and FLOPs are 0.77M and 0.53G, respectively. Experimental results on two widely recognized benchmark datasets demonstrate that MpoxMamba outperforms state-of-the-art lightweight models and existing mpox detection methods. Importantly, we developed a web-based online application to provide free mpox detection (http://5227i971s5.goho.co:30290). The source codes of MpoxMamba are available at https://github.com/YubiaoYue/MpoxMamba.

We construct solitary waves for the fractional Korteweg-De Vries type equation $u_t + (\Lambda^{-s}u + u^2)_x = 0$, where $\Lambda^{-s}$ denotes the Bessel potential operator $(1 + |D|^2)^{-\frac{s}{2}}$ for $s \in (0,\infty)$. The approach is to parameterise the known periodic solution curves through the relative wave height. Using a priori estimates, we show that the periodic waves locally uniformly converge to waves with negative tails, which are transformed to the desired branch of solutions. The obtained branch reaches a highest wave, the behavior of which varies with $s$. The work is a generalisation of recent work by Ehrnstr\"om-Nik-Walker, and is as far as we know the first simultaneous construction of small, intermediate and highest solitary waves for the complete family of (inhomogeneous) fractional KdV equations with negative-order dispersive operators. The obtained waves display exponential decay rate as $|x| \to \infty$., Comment: 21 pages

In the telephony scenarios, the fake speech detection (FSD) task to combat speech spoofing attacks is challenging. Data augmentation (DA) methods are considered effective means to address the FSD task in telephony scenarios, typically divided into time domain and frequency domain stages. While each has its advantages, both can result in information loss. To tackle this issue, we propose a novel DA method, Frequency-mix (Freqmix), and introduce the Freqmix knowledge distillation (FKD) to enhance model information extraction and generalization abilities. Specifically, we use Freqmix-enhanced data as input for the teacher model, while the student model's input undergoes time-domain DA method. We use a multi-level feature distillation approach to restore information and improve the model's generalization capabilities. Our approach achieves state-of-the-art results on ASVspoof 2021 LA dataset, showing a 31\% improvement over baseline and performs competitively on ASVspoof 2021 DF dataset., Comment: Accepted by Interspeech 2024

Fake artefacts for discriminating between bonafide and fake audio can exist in both short- and long-range segments. Therefore, combining local and global feature information can effectively discriminate between bonafide and fake audio. This paper proposes an end-to-end bidirectional state space model, named RawBMamba, to capture both short- and long-range discriminative information for audio deepfake detection. Specifically, we use sinc Layer and multiple convolutional layers to capture short-range features, and then design a bidirectional Mamba to address Mamba's unidirectional modelling problem and further capture long-range feature information. Moreover, we develop a bidirectional fusion module to integrate embeddings, enhancing audio context representation and combining short- and long-range information. The results show that our proposed RawBMamba achieves a 34.1\% improvement over Rawformer on ASVspoof2021 LA dataset, and demonstrates competitive performance on other datasets., Comment: Accepted by Interspeech 2024

In recent years, knowledge graph completion (KGC) models based on pre-trained language model (PLM) have shown promising results. However, the large number of parameters and high computational cost of PLM models pose challenges for their application in downstream tasks. This paper proposes a progressive distillation method based on masked generation features for KGC task, aiming to significantly reduce the complexity of pre-trained models. Specifically, we perform pre-distillation on PLM to obtain high-quality teacher models, and compress the PLM network to obtain multi-grade student models. However, traditional feature distillation suffers from the limitation of having a single representation of information in teacher models. To solve this problem, we propose masked generation of teacher-student features, which contain richer representation information. Furthermore, there is a significant gap in representation ability between teacher and student. Therefore, we design a progressive distillation method to distill student models at each grade level, enabling efficient knowledge transfer from teachers to students. The experimental results demonstrate that the model in the pre-distillation stage surpasses the existing state-of-the-art methods. Furthermore, in the progressive distillation stage, the model significantly reduces the model parameters while maintaining a certain level of performance. Specifically, the model parameters of the lower-grade student model are reduced by 56.7\% compared to the baseline., Comment: Accepted by AAAI2024

The aim of this study is to analyze the effect of serum metabolites on diabetic nephropathy (DN) and predict the prevalence of DN through a machine learning approach. The dataset consists of 548 patients from April 2018 to April 2019 in Second Affiliated Hospital of Dalian Medical University (SAHDMU). We select the optimal 38 features through a Least absolute shrinkage and selection operator (LASSO) regression model and a 10-fold cross-validation. We compare four machine learning algorithms, including eXtreme Gradient Boosting (XGB), random forest, decision tree and logistic regression, by AUC-ROC curves, decision curves, calibration curves. We quantify feature importance and interaction effects in the optimal predictive model by Shapley Additive exPlanations (SHAP) method. The XGB model has the best performance to screen for DN with the highest AUC value of 0.966. The XGB model also gains more clinical net benefits than others and the fitting degree is better. In addition, there are significant interactions between serum metabolites and duration of diabetes. We develop a predictive model by XGB algorithm to screen for DN. C2, C5DC, Tyr, Ser, Met, C24, C4DC, and Cys have great contribution in the model, and can possibly be biomarkers for DN.

Auditory Attention Detection (AAD) aims to detect target speaker from brain signals in a multi-speaker environment. Although EEG-based AAD methods have shown promising results in recent years, current approaches primarily rely on traditional convolutional neural network designed for processing Euclidean data like images. This makes it challenging to handle EEG signals, which possess non-Euclidean characteristics. In order to address this problem, this paper proposes a dynamical graph self-distillation (DGSD) approach for AAD, which does not require speech stimuli as input. Specifically, to effectively represent the non-Euclidean properties of EEG signals, dynamical graph convolutional networks are applied to represent the graph structure of EEG signals, which can also extract crucial features related to auditory spatial attention in EEG signals. In addition, to further improve AAD detection performance, self-distillation, consisting of feature distillation and hierarchical distillation strategies at each layer, is integrated. These strategies leverage features and classification results from the deepest network layers to guide the learning of shallow layers. Our experiments are conducted on two publicly available datasets, KUL and DTU. Under a 1-second time window, we achieve results of 90.0\% and 79.6\% accuracy on KUL and DTU, respectively. We compare our DGSD method with competitive baselines, and the experimental results indicate that the detection performance of our proposed DGSD method is not only superior to the best reproducible baseline but also significantly reduces the number of trainable parameters by approximately 100 times.

Booming deep learning has substantially improved the diagnosis for diverse lesions in ultrasound images, but a conspicuous research gap concerning cervical lymph node lesions still remains. The objective of this work is to diagnose cervical lymph node lesions in ultrasound images by leveraging a deep learning model. To this end, we first collected 3392 cervical ultrasound images containing normal lymph nodes, benign lymph node lesions, malignant primary lymph node lesions, and malignant metastatic lymph node lesions. Given that ultrasound images are generated by the reflection and scattering of sound waves across varied bodily tissues, we proposed the Conv-FFT Block. It integrates convolutional operations with the fast Fourier transform to more astutely model the images. Building upon this foundation, we designed a novel architecture, named SFUSNet. SFUSNet not only discerns variances in ultrasound images from the spatial domain but also adeptly captures micro-structural alterations across various lesions in the frequency domain. To ascertain the potential of SFUSNet, we benchmarked it against 12 popular architectures through five-fold cross-validation. The results show that SFUSNet is the state-of-the-art model and can achieve 92.89% accuracy. Moreover, its average precision, average sensitivity and average specificity for four types of lesions achieve 90.46%, 89.95% and 97.49%, respectively.

Numerous studies have affirmed that deep learning models can facilitate early diagnosis of lesions in endoscopic images. However, the lack of available datasets stymies advancements in research on nasal endoscopy, and existing models fail to strike a good trade-off between model diagnosis performance, model complexity and parameters size, rendering them unsuitable for real-world application. To bridge these gaps, we created the first large-scale nasal endoscopy dataset, named 7-NasalEID, comprising 11,352 images that contain six common nasal diseases and normal samples. Subsequently, we proposed U-SEANNet, an innovative U-shaped architecture, underpinned by depth-wise separable convolution. Moreover, to enhance its capacity for detecting nuanced discrepancies in input images, U-SEANNet employs the Global-Local Channel Feature Fusion module, enabling it to utilize salient channel features from both global and local contexts. To demonstrate U-SEANNet's potential, we benchmarked U-SEANNet against seventeen modern architectures through five-fold cross-validation. The experimental results show that U-SEANNet achieves a commendable accuracy of 93.58%. Notably, U-SEANNet's parameters size and GFLOPs are only 0.78M and 0.21, respectively. Our findings suggest U-SEANNet is the state-of-the-art model for nasal diseases diagnosis in endoscopic images., Comment: There are some descriptive errors in the manuscript

The rhythm of bonafide speech is often difficult to replicate, which causes that the fundamental frequency (F0) of synthetic speech is significantly different from that of real speech. It is expected that the F0 feature contains the discriminative information for the fake speech detection (FSD) task. In this paper, we propose a novel F0 subband for FSD. In addition, to effectively model the F0 subband so as to improve the performance of FSD, the spatial reconstructed local attention Res2Net (SR-LA Res2Net) is proposed. Specifically, Res2Net is used as a backbone network to obtain multiscale information, and enhanced with a spatial reconstruction mechanism to avoid losing important information when the channel group is constantly superimposed. In addition, local attention is designed to make the model focus on the local information of the F0 subband. Experimental results on the ASVspoof 2019 LA dataset show that our proposed method obtains an equal error rate (EER) of 0.47% and a minimum tandem detection cost function (min t-DCF) of 0.0159, achieving the state-of-the-art performance among all of the single systems., Comment: Accept by Neural Networks

In this paper, we propose the multi-perspective information fusion (MPIF) Res2Net with random Specmix for fake speech detection (FSD). The main purpose of this system is to improve the model's ability to learn precise forgery information for FSD task in low-quality scenarios. The task of random Specmix, a data augmentation, is to improve the generalization ability of the model and enhance the model's ability to locate discriminative information. Specmix cuts and pastes the frequency dimension information of the spectrogram in the same batch of samples without introducing other data, which helps the model to locate the really useful information. At the same time, we randomly select samples for augmentation to reduce the impact of data augmentation directly changing all the data. Once the purpose of helping the model to locate information is achieved, it is also important to reduce unnecessary information. The role of MPIF-Res2Net is to reduce redundant interference information. Deceptive information from a single perspective is always similar, so the model learning this similar information will produce redundant spoofing clues and interfere with truly discriminative information. The proposed MPIF-Res2Net fuses information from different perspectives, making the information learned by the model more diverse, thereby reducing the redundancy caused by similar information and avoiding interference with the learning of discriminative information. The results on the ASVspoof 2021 LA dataset demonstrate the effectiveness of our proposed method, achieving EER and min-tDCF of 3.29% and 0.2557, respectively., Comment: Accepted by DADA2023

In this paper, we propose a novel self-distillation method for fake speech detection (FSD), which can significantly improve the performance of FSD without increasing the model complexity. For FSD, some fine-grained information is very important, such as spectrogram defects, mute segments, and so on, which are often perceived by shallow networks. However, shallow networks have much noise, which can not capture this very well. To address this problem, we propose using the deepest network instruct shallow network for enhancing shallow networks. Specifically, the networks of FSD are divided into several segments, the deepest network being used as the teacher model, and all shallow networks become multiple student models by adding classifiers. Meanwhile, the distillation path between the deepest network feature and shallow network features is used to reduce the feature difference. A series of experimental results on the ASVspoof 2019 LA and PA datasets show the effectiveness of the proposed method, with significant improvements compared to the baseline., Comment: Accepted by ICASSP 2023

Recently, pioneer research works have proposed a large number of acoustic features (log power spectrogram, linear frequency cepstral coefficients, constant Q cepstral coefficients, etc.) for audio deepfake detection, obtaining good performance, and showing that different subbands have different contributions to audio deepfake detection. However, this lacks an explanation of the specific information in the subband, and these features also lose information such as phase. Inspired by the mechanism of synthetic speech, the fundamental frequency (F0) information is used to improve the quality of synthetic speech, while the F0 of synthetic speech is still too average, which differs significantly from that of real speech. It is expected that F0 can be used as important information to discriminate between bonafide and fake speech, while this information cannot be used directly due to the irregular distribution of F0. Insteadly, the frequency band containing most of F0 is selected as the input feature. Meanwhile, to make full use of the phase and full-band information, we also propose to use real and imaginary spectrogram features as complementary input features and model the disjoint subbands separately. Finally, the results of F0, real and imaginary spectrogram features are fused. Experimental results on the ASVspoof 2019 LA dataset show that our proposed system is very effective for the audio deepfake detection task, achieving an equivalent error rate (EER) of 0.43%, which surpasses almost all systems.

We prove the existence of highest, cusped, periodic travelling-wave solutions with exact and optimal $ \alpha $-H\"older continuity in a class of fractional negative-order dispersive equations of the form \begin{equation*} u_t + (| \mathrm{D} |^{- \alpha} u + n(u) )_x = 0 \end{equation*} for every $ \alpha \in (0, 1) $ with homogeneous Fourier multiplier $ | \mathrm{D} |^{ - \alpha} $. We tackle nonlinearities $ n(u) $ of the type $ | u |^p $ or $ u | u |^{p - 1} $ for all real $ p > 1 $, and show that when $ n $ is odd, the waves also feature antisymmetry and thus contain inverted cusps. Tools involve detailed pointwise estimates in tandem with analytic global bifurcation, where we resolve the issue with nonsmooth $ n $ by means of regularisation. We believe that both the construction of highest antisymmetric waves and the regularisation of nonsmooth terms to an analytic bifurcation setting are new in this context, with direct applicability also to generalised versions of the Whitham, the Burgers--Poisson, the Burgers--Hilbert, the Degasperis--Procesi, the reduced Ostrovsky, and the bidirectional Whitham equations., Comment: 33 pages; 5 figures

We prove global existence, uniqueness and stability of entropy solutions with $L^2\cap L^\infty$ initial data for a general family of negative order dispersive equations. It is further demonstrated that this solution concept extends in a unique continuous manner to all $L^2$ initial data. These weak solutions are found to satisfy one sided H\"older conditions whose coefficients decay in time. The latter result controls the height of solutions and further provides a way to bound the maximal lifespan of classical solutions from their initial data., Comment: 46 pages

We describe our participation in all five rounds of the TREC 2020 COVID Track (TREC-COVID). The goal of TREC-COVID is to contribute to the response to the COVID-19 pandemic by identifying answers to many pressing questions and building infrastructure to improve search systems [8]. All five rounds of this Track challenged participants to perform a classic ad-hoc search task on the new data collection CORD-19. Our solution addressed this challenge by applying the Continuous Active Learning model (CAL) and its variations. Our results showed us to be amongst the top scoring manual runs and we remained competitive within all categories of submissions., Comment: 7 pages, 5 figures

Investigating and classifying sentiments of social media users towards an item, situation, and system are very popular among the researchers. However, they rarely discuss the underlying socioeconomic factor associations for such sentiments. This study attempts to explore the factors associated with positive and negative sentiments of the people about reopening the economy, in the United States (US) amidst the COVID-19 global crisis. It takes into consideration the situational uncertainties (i.e., changes in work and travel pattern due to lockdown policies), economic downturn and associated trauma, and emotional factors such as depression. To understand the sentiment of the people about the reopening economy, Twitter data was collected, representing the 51 states including Washington DC of the US. State-wide socioeconomic characteristics of the people, built environment data, and the number of COVID-19 related cases were collected and integrated with Twitter data to perform the analysis. A binary logit model was used to identify the factors that influence people toward a positive or negative sentiment. The results from the logit model demonstrate that family households, people with low education levels, people in the labor force, low-income people, and people with higher house rent are more interested in reopening the economy. In contrast, households with a high number of members and high income are less interested to reopen the economy. The model can correctly classify 56.18% of the sentiments. The Pearson chi2 test indicates that overall this model has high goodness-of-fit. This study provides a clear indication to the policymakers where to allocate resources and what policy options they can undertake to improve the socioeconomic situations of the people and mitigate the impacts of pandemics in the current situation and as well as in the future.

In this work, we consider the following generalized derivative nonlinear Schr\"odinger equation \begin{align*} i\partial_t u+\partial_{xx} u +i |u|^{2\sigma}\partial_x u=0, \quad (t,x)\in \mathbb R\times \mathbb R. \end{align*} We prove that when $\sigma\ge 2$, the solution is global and scattering when the initial data is small in $H^s(\mathbb R)$, $\frac 12\leq s\leq1$. Moreover, we show that when $0<\sigma<2$, there exist a class of solitary wave solutions $\{\phi_c\}$ satisfying $$ \|\phi_c\|_{H^1(\mathbb R)}\to 0, $$ when $c$ tends to some endpoint, which is against the small data scattering statement. Therefore, in this model, the exponent $\sigma\ge2$ is optimal for small data scattering. We remark that this exponent is larger than the short range exponent and the Strauss exponent., Comment: 26 pages

In this work, we consider the following generalized Boussinesq equation \begin{align*} \partial_{t}^2u-\partial_{x}^2u+\partial_{x}^2(\partial_{x}^2u+|u|^{p}u)=0,\qquad (t,x)\in\mathbb R\times \mathbb R, \end{align*} with $0

In the past two decades, under the conditions that both the osmotic pressure $\Delta p$ and tensile stress $\lambda$ equal zero, a rigorous solution (RS) of human red blood cell (RBC) with a minus spontaneous curvature $c_{0}$ has been derived with Helfrich model. And, by fitting with observed shapes of RBC, $c_{0}R_{0}$ has been predicted to be -1.62 as minus golden ratio, where $R_{0}$ is the radius of a sphere with the same area of RBC. In this Lett., it is also found $\rho_{max}$ /$\rho_{B}\approx$ 1.6 shows a approximately beautiful golden cross section of RBC, where $\rho_{max}$ is the radius of RBC and $\rho_{B}$ is the radius at maximal thickness of RBC. With a complete numerical calculation, we find the mechanism behind the beauty that minus golden ratio of $c_{0}R_{0}$ is the balance between economical surface area and enough deformability to pass spleen, the so called "physical fitness test"., Comment: 4pages, 5figures

The CSR-LINAC injector has been proposed in Heavy Ion Research Facility in Lanzhou (HIRFL). The linac mainly consists of two parts, the RFQ and the IH-DTL. The KONUS (Kombinierte Null Grad Struktur) concept has been introduced into the DTL section. In this paper, the re-matching of the main accelerating section will be finished in the 3.7 MeV/u scheme and the new beam dynamics design up to 7 MeV/u will be also shown. Through the beam re-matching, the relative emittance growth has been suppressed greatly along the linac.

In this paper, we extend the first-order post-Newtonian scheme in multiple systems presented by Damour-Soffel-Xu to the second-order contribution to light propagation without changing the virtueof the scheme on the linear partial differential equations of the potential and vector potential. The spatial components of the metric are extended to second order level both in a global coordinates ($q_{ij}/ c^4$) and a local coordinates ($Q_{ab}/ c^4$). The equations of $q_{ij}$ (or $Q_{ab}$) are obtained from the field equations.The relationship between $q_{ij}$ and $Q_{ab}$ are presented in this paper also. In special case of the solar system (isotropic condition is applied ($q_{ij} = \delta_{ij} q $)), we obtain the solution of $q$. Finally, a further extension of the second-order contributions in the parametrized post-Newtonian formalism is discussed., Comment: Latex2e; 6 pages PS file