arXiv subjects
Publications and source records attributed to Hui Liu.
We scatter a meson off of a scalar kink in quantum field theory, at leading order in perturbation theory. We calculate the full quantum state, at leading order, at all times and also check that the reflection and transmission coefficients agree with those which would be obtained in relativistic quantum mechanics.
Sarcasm is a linguistic phenomenon indicating a discrepancy between literal meanings and implied intentions. Due to its sophisticated nature, it is usually challenging to be detected from the text itself. As a result, multi-modal sarcasm detection has received more attention in both academia and industries. However, most existing techniques only modeled the atomic-level inconsistencies between the text input and its accompanying image, ignoring more complex compositions for both modalities. Moreover, they neglected the rich information contained in external knowledge, e.g., image captions. In this paper, we propose a novel hierarchical framework for sarcasm detection by exploring both the atomic-level congruity based on multi-head cross attention mechanism and the composition-level congruity based on graph neural networks, where a post with low congruity can be identified as sarcasm. In addition, we exploit the effect of various knowledge resources for sarcasm detection. Evaluation results on a public multi-modal sarcasm detection dataset based on Twitter demonstrate the superiority of our proposed model.
Pseudo-Goldstone modes appear in many physical systems and display robust universal features. First, their mass $m$ obeys the so-called Gell-Mann-Oakes-Renner (GMOR) relation $f^2\,m^2=H\,\barσ$, with $f$ the Goldstone stiffness, $H$ the explicit breaking scale and $\barσ$ the spontaneous condensate. More recently, it has been shown that their damping $Ω$ is constrained to follow the relation $Ω=m^2 D_φ$, where $D_φ$ is the Goldstone diffusivity in the purely spontaneous phase. Pions are the most paradigmatic example of pseudo-Goldstone modes and they are related to chiral symmetry breaking in QCD. In this work, we consider a bottom-up soft-wall AdS-QCD model with broken ${\rm{SU}}(2)_L \times {\rm{SU}}(2)_R$ symmetry and we study the nature of the associated pseudo-Goldstone modes -- the pions. In particular, we perform a detailed investigation of their dispersion relation in presence of dissipation, of the role of the explicit breaking induced by the quark masses and of the dynamics near the critical point. Taking advantage of the microscopic information provided by the holographic model, we give quantitative predictions for all the coefficients appearing in the effective description. In particular, we estimate the finite temperature behavior of the kinetic parameter $\mathfrak{r^2}$ defined as the ration between the Goldstone diffusivity $D_φ$ and the pion attenuation constant $D_A$. Interestingly, we observe important deviations from the value $\mathfrak{r^2}=3/4$ computed in chiral perturbation theory in the limit of zero temperature.
Although deep neural networks (DNNs) have shown impressive performance on many perceptual tasks, they are vulnerable to adversarial examples that are generated by adding slight but maliciously crafted perturbations to benign images. Adversarial detection is an important technique for identifying adversarial examples before they are entered into target DNNs. Previous studies to detect adversarial examples either targeted specific attacks or required expensive computation. How design a lightweight unsupervised detector is still a challenging problem. In this paper, we propose an AutoEncoder-based Adversarial Examples (AEAE) detector, that can guard DNN models by detecting adversarial examples with low computation in an unsupervised manner. The AEAE includes only a shallow autoencoder but plays two roles. First, a well-trained autoencoder has learned the manifold of benign examples. This autoencoder can produce a large reconstruction error for adversarial images with large perturbations, so we can detect significantly perturbed adversarial examples based on the reconstruction error. Second, the autoencoder can filter out the small noise and change the DNN's prediction on adversarial examples with small perturbations. It helps to detect slightly perturbed adversarial examples based on the prediction distance. To cover these two cases, we utilize the reconstruction error and prediction distance from benign images to construct a two-tuple feature set and train an adversarial detector using the isolation forest algorithm. We show empirically that the AEAE is unsupervised and inexpensive against the most state-of-the-art attacks. Through the detection in these two cases, there is nowhere to hide adversarial examples.
Let $X=\{X_n: n\in\mathbb{N}\}$ be a long memory linear process with innovations in the domain of attraction of an $α$-stable law $(0<α<2)$. Assume that the linear process $X$ has a bounded probability density function $f(x)$. Then, under certain conditions, we consider the estimation of the quadratic functional $\int_{\mathbb{R}} f^2(x) \,dx$ by using the kernel estimator \[ T_n(h_n)=\frac{2}{n(n-1)h_n}\sum_{1\leq j<i\leq n}K\left(\frac{X_i-X_j}{h_n}\right). \] The simulation study for long memory linear processes with symmetric $α$-stable innovations is also given.
Millimeter-wave (mmW) radars are being increasingly integrated in commercial vehicles to support new Adaptive Driver Assisted Systems (ADAS) for its ability to provide high accuracy location, velocity, and angle estimates of objects, largely independent of environmental conditions. Such radar sensors not only perform basic functions such as detection and ranging/angular localization, but also provide critical inputs for environmental perception via object recognition and classification. To explore radar-based ADAS applications, we have assembled a lab-scale frequency modulated continuous wave (FMCW) radar test-bed (https://depts.washington.edu/funlab/research) based on Texas Instrument's (TI) automotive chipset family. In this work, we describe the test-bed components and provide a summary of FMCW radar operational principles. To date, we have created a large raw radar dataset for various objects under controlled scenarios. Thereafter, we apply some radar imaging algorithms to the collected dataset, and present some preliminary results that validate its capabilities in terms of object recognition. Our code is available at https://github.com/Xiangyu-Gao/mmWave-radar-signal-processing-and-microDoppler-classification.
Light scattering within scattering media presents a substantial obstacle to optical transmission. A speckle pattern with random amplitude and phase distribution is observed when coherent light travels through strong scattering media. Fortunately, wavefront shaping has been successfully employed with a spatial light modulator to recover intensity targets after scattering media, such as a sharp focus point or specified two-dimensional patterns. There have, however, been few studies that attempted to separately manipulate the amplitude and phase of the focusing field. In this paper, we propose a feedback-based wavefront shaping method to generate complex optical fields through scattering medium. A reliable phase retrieval approach is introduced to provide the complex feedback information, i.e., the amplitude and phase of the focusing field. Accordingly, in order to modulate the speckle field into a desired complex structured optical field, a multi-objective genetic algorithm is used to find the best phase map. To demonstrate the proposed method's high performance, experimental tests have been carried out. High fidelity is demonstrated in the generation of diverse complex light fields, both in amplitude and phase. Our findings may facilitate the manipulation of light field through scattering medium, and are anticipated to further promote future applications such as optogenetics, vortex optical communication, and optical trapping through scattering media.
In these proceedings, we present the measurements of centrality, transverse momentum and rapidity dependences of proton ($p$) and light nuclei ($d$ ($\overline{d}$), $t$, $^{3}\mathrm{He}$ ($\overline{^{3}\mathrm{He}}$), and $^{4}\mathrm{He}$) production in Au+Au collisions at $\sqrt{s_{\mathrm{NN}}}$ = 3 GeV, and isobaric (Ru+Ru and Zr+Zr) collisions at $\sqrt{s_{\mathrm{NN}}}$ = 200 GeV. The compound yield ratios in central collisions at 3 GeV are found to be larger than the transport model calculations. Furthermore, the kinetic freeze-out parameters at 3 GeV show a different trend compared to those of light hadrons ($π$, $K$, $p$) at higher energies.
Mainstream object detectors are commonly constituted of two sub-tasks, including classification and regression tasks, implemented by two parallel heads. This classic design paradigm inevitably leads to inconsistent spatial distributions between classification score and localization quality (IOU). Therefore, this paper alleviates this misalignment in the view of knowledge distillation. First, we observe that the massive teacher achieves a higher proportion of harmonious predictions than the lightweight student. Based on this intriguing observation, a novel Harmony Score (HS) is devised to estimate the alignment of classification and regression qualities. HS models the relationship between two sub-tasks and is seen as prior knowledge to promote harmonious predictions for the student. Second, this spatial misalignment will result in inharmonious region selection when distilling features. To alleviate this problem, a novel Task-decoupled Feature Distillation (TFD) is proposed by flexibly balancing the contributions of classification and regression tasks. Eventually, HD and TFD constitute the proposed method, named Task-Balanced Distillation (TBD). Extensive experiments demonstrate the considerable potential and generalization of the proposed method. Specifically, when equipped with TBD, RetinaNet with ResNet-50 achieves 41.0 mAP under the COCO benchmark, outperforming the recent FGD and FRS.
We report the synthesis and characterization of a Si-based ternary semiconductor Mg$_3$Si$_2$Te$_6$, which exhibits a quasi-two-dimensional structure, where the trigonal Mg$_2$Si$_2$Te$_6$ layers are separated by Mg ions. Ultraviolet-visible absorption spectroscopy and density functional theory calculations were performed to investigate the electronic structure. The experimentally determined direct band gap is 1.39 eV, consistent with the value of the density function theory calculations. Our results reveal that Mg$_3$Si$_2$Te$_6$ is a direct gap semiconductor with a relatively narrow gap, which is a potential candidate for infrared optoelectronic devices.
This paper tackles the challenging problem of hyperspectral (HS) image denoising. Unlike existing deep learning-based methods usually adopting complicated network architectures or empirically stacking off-the-shelf modules to pursue performance improvement, we focus on the efficient and effective feature extraction manner for capturing the high-dimensional characteristics of HS images. To be specific, based on the theoretical analysis that increasing the rank of the matrix formed by the unfolded convolutional kernels can promote feature diversity, we propose rank-enhanced low-dimensional convolution set (Re-ConvSet), which separately performs 1-D convolution along the three dimensions of an HS image side-by-side, and then aggregates the resulting spatial-spectral embeddings via a learnable compression layer. Re-ConvSet not only learns the diverse spatial-spectral features of HS images, but also reduces the parameters and complexity of the network. We then incorporate Re-ConvSet into the widely-used U-Net architecture to construct an HS image denoising method. Surprisingly, we observe such a concise framework outperforms the most recent method to a large extent in terms of quantitative metrics, visual results, and efficiency. We believe our work may shed light on deep learning-based HS image processing and analysis.
We report high pressure studies on the C-type antiferromagnetic semiconductor EuTe2 up to 36.0 GPa. A structural transition from the I4/mcm to C2/m space group is identified at ~16 GPa. Superconductivity is discovered above ~5 GPa in both the I4/mcm and C2/m space groups. In the low-pressure phase (< 16 GPa), the antiferromagnetic transition temperature is enhanced with increasing pressure due to the enhanced magnetic exchange interactions. Magnetoresistance measurements indicate an interplay between the local moments of Eu2+ and the conduction electrons of Te 5p orbits. The upper critical field of the superconductivity is well above the Pauli limit. Across the structural transition to the high-pressure phase (> 16 GPa), EuTe2 becomes nonmagnetic and the superconducting transition temperature evolves smoothly with the upper critical field below the Pauli limit. Therefore, the high upper critical field of EuTe2 in the low-pressure phase is due to the exchange field compensation effect of the Eu magnetic order and the superconductivity in both structures may arise in the framework of the BCS theory.
A general spherical harmonics method is described for extracting anisotropic pair distribution functions (PDF) in this work. In the structural study of functional crystallized materials, the investigation of the local structures under the application of external stimuli, such as electric field and stress, is in urgent need. A well-established technique for local structure studies is PDF analysis, but the extraction of the X-ray PDF data is usually based on angular integrations of isotropic X-ray structure functions, which is no longer valid for the anisotropic responses of the materials under orientation-dependent stimuli. Therefore, we have developed an advanced spherical harmonics method to transform two-dimensional X-ray diffraction patterns into anisotropic PDF data, based on three-dimensional diffraction geometry and Fourier transform. The electrical-field-induced local structural change in the PbZr0.54Ti0.46O3 ceramics is then presented to demonstrate the method's effectiveness.
Suppressing the interference of atmospheric turbulence and obtaining observation data with a high spatial resolution is an issue to be solved urgently for ground observations. One way to solve this problem is to perform a statistical reconstruction of short-exposure speckle images. Combining the rapidity of Shift-Add and the accuracy of speckle masking, this paper proposes a novel reconstruction algorithm-NASIR (Non-rigid Alignment based Solar Image Reconstruction). NASIR reconstructs the phase of the object image at each frequency by building a computational model between geometric distortion and intensity distribution and reconstructs the modulus of the object image on the aligned speckle images by speckle interferometry. We analyzed the performance of NASIR by using the correlation coefficient, power spectrum, and coefficient of variation of intensity profile (CVoIP) in processing data obtained by the NVST (1m New Vacuum Solar Telescope). The reconstruction experiments and analysis results show that the quality of images reconstructed by NASIR is close to speckle masking when the seeing is good, while NASIR has excellent robustness when the seeing condition becomes worse. Furthermore, NASIR reconstructs the entire field of view in parallel in one go, without phase recursion and block-by-block reconstruction, so its computation time is less than half that of speckle masking. Therefore, we consider NASIR is a robust and high-quality fast reconstruction method that can serve as an effective tool for data filtering and quick look.
Handwritten mathematical expression recognition (HMER) is a challenging task that has many potential applications. Recent methods for HMER have achieved outstanding performance with an encoder-decoder architecture. However, these methods adhere to the paradigm that the prediction is made "from one character to another", which inevitably yields prediction errors due to the complicated structures of mathematical expressions or crabbed handwritings. In this paper, we propose a simple and efficient method for HMER, which is the first to incorporate syntax information into an encoder-decoder network. Specifically, we present a set of grammar rules for converting the LaTeX markup sequence of each expression into a parsing tree; then, we model the markup sequence prediction as a tree traverse process with a deep neural network. In this way, the proposed method can effectively describe the syntax context of expressions, alleviating the structure prediction errors of HMER. Experiments on three benchmark datasets demonstrate that our method achieves better recognition performance than prior arts. To further validate the effectiveness of our method, we create a large-scale dataset consisting of 100k handwritten mathematical expression images acquired from ten thousand writers. The source code, new dataset, and pre-trained models of this work will be publicly available.
In this paper, we prove that if all closed characteristics on a compact non-degenerate star-shaped hypersurface $Σ$ in $\mathbf{R}^{2n}$ are elliptic, then either there exist exactly $n$ geometrically distinct closed characteristics, or there exist infinitely many geometrically distinct closed characteristics.
Deep neural networks (DNNs) are threatened by adversarial examples. Adversarial detection, which distinguishes adversarial images from benign images, is fundamental for robust DNN-based services. Image transformation is one of the most effective approaches to detect adversarial examples. During the last few years, a variety of image transformations have been studied and discussed to design reliable adversarial detectors. In this paper, we systematically synthesize the recent progress on adversarial detection via image transformations with a novel classification method. Then, we conduct extensive experiments to test the detection performance of image transformations against state-of-the-art adversarial attacks. Furthermore, we reveal that each individual transformation is not capable of detecting adversarial examples in a robust way, and propose a DNN-based approach referred to as \emph{AdvJudge}, which combines scores of 9 image transformations. Without knowing which individual scores are misleading or not misleading, AdvJudge can make the right judgment, and achieve a significant improvement in detection rate. Finally, we utilize an explainable AI tool to show the contribution of each image transformation to adversarial detection. Experimental results show that the contribution of image transformations to adversarial detection is significantly different, the combination of them can significantly improve the generic detection ability against state-of-the-art adversarial attacks.
It has been verified that only a small fraction of the neoantigens presented by MHC class I molecules on the cell surface can elicit T cells. The limitation can be attributed to the binding specificity of T cell receptor (TCR) to peptide-MHC complex (pMHC). Computational prediction of T cell binding to neoantigens is an challenging and unresolved task. In this paper, we propose an attentive-mask contrastive learning model, ATMTCR, for inferring TCR-antigen binding specificity. For each input TCR sequence, we used Transformer encoder to transform it to latent representation, and then masked a proportion of residues guided by attention weights to generate its contrastive view. Pretraining on large-scale TCR CDR3 sequences, we verified that contrastive learning significantly improved the prediction performance of TCR binding to peptide-MHC complex (pMHC). Beyond the detection of important amino acids and their locations in the TCR sequence, our model can also extracted high-order semantic information underlying the TCR-antigen binding specificity. Comparison experiments were conducted on two independent datasets, our method achieved better performance than other existing algorithms. Moreover, we effectively identified important amino acids and their positional preferences through attention weights, which indicated the interpretability of our proposed model.