arXiv subjects

Yu Zhang

Publications and source records attributed to Yu Zhang.

At least 1,117 records · Page 62Linked to original sources

arXiv2022

Localization, multifractality, and many-body localization in periodically kicked quasiperiodic lattices

We study the combined effect of quasiperiodic disorder, driven and interaction in the periodically kicked Aubry-André model. In the non-interacting limit, by analyzing the quasienergy spectrum statistics, we verify the existence of a dynamical localization transition in the high-frequency region, whereas the spectrum statistics becomes intricate in the low-frequency region due to the emergence of the extended/localized-to-multifractal edges in the quasienergy spectrum, which separate the multifractal states from the extended (localized) states. When the interaction is introduced, we find the periodically kicked incommensurate potential can lead to a transition from ergodic to many-body-localization phase in the high-frequency region. However, the many-body localization phase vanishes in the low-frequency region even for strong quasiperiodic disorder. Our studies demonstrate that the periodically kicked Aubry-André model displays rich dynamical phenomena and the driving frequency plays an important role in the formation of many-body localization in addition to the disorder strength.

cond-mat.dis-nn↗

arXiv2022

See What You See: Self-supervised Cross-modal Retrieval of Visual Stimuli from Brain Activity

Recent studies demonstrate the use of a two-stage supervised framework to generate images that depict human perception to visual stimuli from EEG, referring to EEG-visual reconstruction. They are, however, unable to reproduce the exact visual stimulus, since it is the human-specified annotation of images, not their data, that determines what the synthesized images are. Moreover, synthesized images often suffer from noisy EEG encodings and unstable training of generative models, making them hard to recognize. Instead, we present a single-stage EEG-visual retrieval paradigm where data of two modalities are correlated, as opposed to their annotations, allowing us to recover the exact visual stimulus for an EEG clip. We maximize the mutual information between the EEG encoding and associated visual stimulus through optimization of a contrastive self-supervised objective, leading to two additional benefits. One, it enables EEG encodings to handle visual classes beyond seen ones during training, since learning is not directed at class annotations. In addition, the model is no longer required to generate every detail of the visual stimulus, but rather focuses on cross-modal alignment and retrieves images at the instance level, ensuring distinguishable model output. Empirical studies are conducted on the largest single-subject EEG dataset that measures brain activities evoked by image stimuli. We demonstrate the proposed approach completes an instance-level EEG-visual retrieval task which existing methods cannot. We also examine the implications of a range of EEG and visual encoder structures. Furthermore, for a mostly studied semantic-level EEG-visual classification task, despite not using class annotations, the proposed method outperforms state-of-the-art supervised EEG-visual reconstruction approaches, particularly on the capability of open class recognition.

cs.MM↗

arXiv2022

Enhance Accuracy: Sensitivity and Uncertainty Theory in LiDAR Odometry and Mapping

Currently, the improvement of LiDAR poses estimation accuracy is an urgent need for mobile robots. Research indicates that diverse LiDAR points have different influences on the accuracy of pose estimation. This study aimed to select a good point set to enhance accuracy. Accordingly, the sensitivity and uncertainty of LiDAR point residuals were formulated as a fundamental basis for derivation and analysis. High-sensitivity and low -uncertainty point residual terms are preferred to achieve higher pose estimation accuracy. The proposed selection method has been theoretically proven to be capable of achieving a global statistical optimum. It was tested on artificial data and compared with the KITTI benchmark. It was also implemented in LiDAR odometry (LO) and LiDAR inertial odometry (LIO), both indoors and outdoors. The experiments revealed that utilizing selected LiDAR point residuals simultaneously enhances optimization accuracy, decreases residual terms, and guarantees real-time performance.

cs.RO↗

arXiv2022

Higher-order Proton Cumulants in Au+Au Collisions at $\sqrt{s_{\rm NN}}$ = 3 GeV from RHIC-STAR

In these proceedings, we present the higher-order cumulants of proton multiplicity distributions of the fixed-target (FXT) run in Au+Au collisions at $\sqrt{s_{\rm NN}}$ = 3.0 GeV. The cumulant ratios are presented as a function of centrality and collision energy. The proton cumulant ratio C4/C2 is consistent with fluctuations driven by baryon number conservation and indicates an energy regime dominated by hadronic interactions. These data imply that the QCD critical point could exist at energies higher than 3 GeV if created in heavy-ion collisions.

nucl-ex↗

arXiv2022

A Low-speed Intruder Star in Hyades: A Temporary Residence

We hereby report a low-speed (about~21~km$\cdot$~s$^{-1}$ with respect to the Sun) intruder member in the Hyades cluster based on the data in the literature. The results show that the star is a non-native member star for the Hyades, with its radial velocity being smaller than the radial velocity of the Hyades cluster, even exceeding the standard deviation of the radial velocity of the cluster by a factor of 9. Furthermore, by analyzing and comparing the orbits of this star and its host, it may have intruded into its host in the past 2~Myr. If the star's current motion orbit remains unchanged, it may leave its host in the next 2~Myr. This implies that the intruder star may be temporarily residing in the cluster. This study presents the first observational evidence of a star intrusion into a cluster, which suggests that more evidence may be found.

astro-ph.SR↗

arXiv2022

A Study of Modeling Rising Intonation in Cantonese Neural Speech Synthesis

In human speech, the attitude of a speaker cannot be fully expressed only by the textual content. It has to come along with the intonation. Declarative questions are commonly used in daily Cantonese conversations, and they are usually uttered with rising intonation. Vanilla neural text-to-speech (TTS) systems are not capable of synthesizing rising intonation for these sentences due to the loss of semantic information. Though it has become more common to complement the systems with extra language models, their performance in modeling rising intonation is not well studied. In this paper, we propose to complement the Cantonese TTS model with a BERT-based statement/question classifier. We design different training strategies and compare their performance. We conduct our experiments on a Cantonese corpus named CanTTS. Empirical results show that the separate training approach obtains the best generalization performance and feasibility.

eess.AS↗

arXiv2022

Chiral SO(4) spin-valley density wave and degenerate topological superconductivity in magic-angle-twisted bilayer-graphene

Starting from a realistic extended Hubbard model for a $p_{x,y}$-orbital tight-binding model on the Honeycomb lattice, we perform a thorough investigation on the possible electron instabilities in the MA-TBG near the van Hove (VH) dopings. Here we focus on the interplay between the approximate SU(2)$\times$SU(2) symmetry and the $D_3$ symmetry, which leads to intriguing quantum states relevant to recent experiments, as revealed by our systematic RPA based calculations followed by a succeeding mean-field energy minimization for the ground state energy. At the SU(2)$\times$SU(2) symmetric point, the degenerate inter-valley SDW and VDW are mixed into a new state of matter dubbed as the chiral SO(4) spin-valley DW. This state simultaneously hosts three 4-component vectorial spin-valley DW orders with each adopting one wave vector, and the polarization directions of the three DW orders are mutually perpendicular to one another. %in the $\mathbb{R}^4$ space. In the presence of a tiny inter-valley exchange interaction with coefficient $J_H\to 0^{-}$ which breaks the SU(2)$\times$SU(2) symmetry, a pure chiral SDW state is obtained. In the case of $J_H\to 0^{+}$, a nematic VDW+SDW state emerges which possesses a stripy distribution of the charge density, consistent with the recent STM observations. On the aspect of SC, while the triplet $p+ip$ and singlet $d+id$ topological SCs are degenerate at $J_H=0$ near the VH dopings, the former (latter) is favored for $J_H\to 0^{-}$ ($J_H\to 0^{+}$). In addition, the two asymmetric doping-dependent behaviors of the obtained pairing phase diagram are well consistent with experiments.

cond-mat.str-el↗

arXiv2022

Reasonable Effectiveness of Random Weighting: A Litmus Test for Multi-Task Learning

Multi-Task Learning (MTL) has achieved success in various fields. However, how to balance different tasks to achieve good performance is a key problem. To achieve the task balancing, there are many works to carefully design dynamical loss/gradient weighting strategies but the basic random experiments are ignored to examine their effectiveness. In this paper, we propose the Random Weighting (RW) methods, including Random Loss Weighting (RLW) and Random Gradient Weighting (RGW), where an MTL model is trained with random loss/gradient weights sampled from a distribution. To show the effectiveness and necessity of RW methods, theoretically we analyze the convergence of RW and reveal that RW has a higher probability to escape local minima, resulting in better generalization ability. Empirically, we extensively evaluate the proposed RW methods to compare with twelve state-of-the-art methods on five image datasets and two multilingual problems from the XTREME benchmark to show RW methods can achieve comparable performance with state-of-the-art baselines. Therefore, we think that the RW methods are important baselines for MTL and should attract more attentions.

cs.LG↗

arXiv2022

Improving Distortion Robustness of Self-supervised Speech Processing Tasks with Domain Adaptation

Speech distortions are a long-standing problem that degrades the performance of supervisely trained speech processing models. It is high time that we enhance the robustness of speech processing models to obtain good performance when encountering speech distortions while not hurting the original performance on clean speech. In this work, we propose to improve the robustness of speech processing models by domain adversarial training (DAT). We conducted experiments based on the SUPERB framework on five different speech processing tasks. In case we do not always have knowledge of the distortion types for speech data, we analyzed the binary-domain and multi-domain settings, where the former treats all distorted speech as one domain, and the latter views different distortions as different domains. In contrast to supervised training methods, we obtained promising results in target domains where speech data is distorted with different distortions including new unseen distortions introduced during testing.

cs.SD↗

arXiv2022

Higher-order topological states in photonic Thue-Morse quasicrystals: quadrupole insulator and a new origin of corner states

Corner states (CSs) in higher-order topological insulators (HOTIs) have recently been of great interest in both crystals and quasicrystals. In contrast to electronic systems, HOTIs have not been found in photonic quasicrystals (PQCs). Here, we systemically study the higher-order topology in the two-dimensional Thue-Morse photonic quasicrystals (TM-PQCs). Not only the topological phase transition and the non-trivial CSs with fractional charge induced by multipole moments, but also a new type of CSs are found due to the complex structure of TM-PQCs near corners. The different origins of these CSs are also analyzed based on the tight-binding model. Our work opens the door to explore richer HOT physics beyond photonic crystals and the robustness of CSs in PQC shows the potential for applications.

physics.optics↗

arXiv2022

BigSSL: Exploring the Frontier of Large-Scale Semi-Supervised Learning for Automatic Speech Recognition

We summarize the results of a host of efforts using giant automatic speech recognition (ASR) models pre-trained using large, diverse unlabeled datasets containing approximately a million hours of audio. We find that the combination of pre-training, self-training and scaling up model size greatly increases data efficiency, even for extremely large tasks with tens of thousands of hours of labeled data. In particular, on an ASR task with 34k hours of labeled data, by fine-tuning an 8 billion parameter pre-trained Conformer model we can match state-of-the-art (SoTA) performance with only 3% of the training data and significantly improve SoTA with the full training set. We also report on the universal benefits gained from using big pre-trained and self-trained models for a large set of downstream tasks that cover a wide range of speech domains and span multiple orders of magnitudes of dataset sizes, including obtaining SoTA performance on many public benchmarks. In addition, we utilize the learned representation of pre-trained networks to achieve SoTA results on non-ASR tasks.

eess.AS↗

arXiv2022

Pushing the Limits of Semi-Supervised Learning for Automatic Speech Recognition

We employ a combination of recent developments in semi-supervised learning for automatic speech recognition to obtain state-of-the-art results on LibriSpeech utilizing the unlabeled audio of the Libri-Light dataset. More precisely, we carry out noisy student training with SpecAugment using giant Conformer models pre-trained using wav2vec 2.0 pre-training. By doing so, we are able to achieve word-error-rates (WERs) 1.4%/2.6% on the LibriSpeech test/test-other sets against the current state-of-the-art WERs 1.7%/3.3%.

eess.AS↗

arXiv2022

Dense Cross-Query-and-Support Attention Weighted Mask Aggregation for Few-Shot Segmentation

Research into Few-shot Semantic Segmentation (FSS) has attracted great attention, with the goal to segment target objects in a query image given only a few annotated support images of the target class. A key to this challenging task is to fully utilize the information in the support images by exploiting fine-grained correlations between the query and support images. However, most existing approaches either compressed the support information into a few class-wise prototypes, or used partial support information (e.g., only foreground) at the pixel level, causing non-negligible information loss. In this paper, we propose Dense pixel-wise Cross-query-and-support Attention weighted Mask Aggregation (DCAMA), where both foreground and background support information are fully exploited via multi-level pixel-wise correlations between paired query and support features. Implemented with the scaled dot-product attention in the Transformer architecture, DCAMA treats every query pixel as a token, computes its similarities with all support pixels, and predicts its segmentation label as an additive aggregation of all the support pixels' labels -- weighted by the similarities. Based on the unique formulation of DCAMA, we further propose efficient and effective one-pass inference for n-shot segmentation, where pixels of all support images are collected for the mask aggregation at once. Experiments show that our DCAMA significantly advances the state of the art on standard FSS benchmarks of PASCAL-5i, COCO-20i, and FSS-1000, e.g., with 3.1%, 9.7%, and 3.6% absolute improvements in 1-shot mIoU over previous best records. Ablative studies also verify the design DCAMA.

cs.CV↗

arXiv2022

Optical Observations of the Nearby Type Ia Supernova 2021hpr

We present the optical photometric and spectroscopic observations of the nearby Type Ia supernova (SN) 2021hpr. The observations covered the phase of $-$14.37 to +63.68 days relative to its maximum luminosity in the $B$ band. The evolution of multiband light/color curves of SN 2021hpr is similar to that of normal Type Ia supernovae (SNe Ia) with the exception of some phases, especially a plateau phase that appeared in the $V-R$ color curve before peak luminosity, which resembles that of SN 2017cbv. The first spectrum we observed at t $\sim -$14.4 days shows a higher velocity for the Si II $λ$6355 feature ($\sim$ 21,000 km s$^{-1}$) than that of other normal Velocity (NV) SNe Ia at the same phase. Based on the Si II $λ$6355 velocity of $\sim$ 12,420 km s$^{-1}$ around the maximum light, we deduce that SN 2021hpr is a transitional object between high velocity (HV) and NV SNe Ia. Meanwhile, the Si II $λ$6355 feature shows a high velocity gradient (HVG) of about 800 km s$^{-1}$ day$^{-1}$ from roughly $-$14.37 to $-$4.31 days relative to the $B$-band maximum, which indicates that SN 2021hpr can also be classified as an HVG SN Ia. The evolution of SN 2021hpr is similar to that of SN 2011fe. Including SN 2021hpr, there have been six supernovae observed in the host galaxy NGC 3147, and the supernovae explosion rate in the last 50 yr is slightly higher for SNe Ia, while lower for SNe Ibc and SNe II it is lower than expected rate from the radio data. Inspecting the spectra, we find that SN 2021hpr has a metal-rich (12 + log(O/H) $\approx$ 8.648) circumstellar environment, where HV SNe tend to reside. Based on the decline rate of SN 2021hpr in the $B$ band, we determine the distance modulus of the host galaxy NGC 3147 using the Phillips relation to be 33.46 $\pm$ 0.21 mag, which is close to that found by previous works.

astro-ph.HE↗

arXiv2022

Contrastive Graph Learning for Population-based fMRI Classification

Contrastive self-supervised learning has recently benefited fMRI classification with inductive biases. Its weak label reliance prevents overfitting on small medical datasets and tackles the high intraclass variances. Nonetheless, existing contrastive methods generate resemblant pairs only on pixel-level features of 3D medical images, while the functional connectivity that reveals critical cognitive information is under-explored. Additionally, existing methods predict labels on individual contrastive representation without recognizing neighbouring information in the patient group, whereas interpatient contrast can act as a similarity measure suitable for population-based classification. We hereby proposed contrastive functional connectivity graph learning for population-based fMRI classification. Representations on the functional connectivity graphs are "repelled" for heterogeneous patient pairs meanwhile homogeneous pairs "attract" each other. Then a dynamic population graph that strengthens the connections between similar patients is updated for classification. Experiments on a multi-site dataset ADHD200 validate the superiority of the proposed method on various metrics. We initially visualize the population relationships and exploit potential subtypes.

cs.LG↗

arXiv2022

Masked Spatial-Spectral Autoencoders Are Excellent Hyperspectral Defenders

Deep learning methodology contributes a lot to the development of hyperspectral image (HSI) analysis community. However, it also makes HSI analysis systems vulnerable to adversarial attacks. To this end, we propose a masked spatial-spectral autoencoder (MSSA) in this paper under self-supervised learning theory, for enhancing the robustness of HSI analysis systems. First, a masked sequence attention learning module is conducted to promote the inherent robustness of HSI analysis systems along spectral channel. Then, we develop a graph convolutional network with learnable graph structure to establish global pixel-wise combinations.In this way, the attack effect would be dispersed by all the related pixels among each combination, and a better defense performance is achievable in spatial aspect.Finally, to improve the defense transferability and address the problem of limited labelled samples, MSSA employs spectra reconstruction as a pretext task and fits the datasets in a self-supervised manner.Comprehensive experiments over three benchmarks verify the effectiveness of MSSA in comparison with the state-of-the-art hyperspectral classification methods and representative adversarial defense strategies.

cs.CV↗

arXiv2022

Entanglement Dynamics in Anti-$\mathcal{PT}$-Symmetric Systems

In the past years, many efforts have been made to study various noteworthy phenomena in both parity-time ($\mathcal{PT}$) and anti-parity-time ($\mathcal{APT}$) symmetric systems. However, entanglement dynamics in $\mathcal{APT}$-symmetric systems has not previously been investigated in both theory and experiments. Here, we investigate the entanglement evolution of two qubits in an $\mathcal{APT}$-symmetric system. In the $\mathcal{APT}$-symmetric unbroken regime, our theoretical simulations demonstrate the periodic oscillations of entanglement when each qubit evolves identically, while the nonperiodic oscillations of entanglement when each qubit evolves differently. In particular, when each qubit evolves near the exceptional point in the $\mathcal{APT}$-symmetric unbroken regime, there exist entanglement sudden vanishing and revival. Moreover, our simulations demonstrate rapid decay and delayed death of entanglement provided one qubit evolves in the $\mathcal{APT}$-symmetric broken regime. In this work, we also perform an experiment with a linear optical setup. The experimental results agree well with our theoretical simulation results. Our findings reveal novel phenomena of entanglement evolution in the $\mathcal{APT}$-symmetric system and opens a new direction for future studies on the dynamics of quantum entanglement in multiqubit $\mathcal{APT}$-symmetric systems or other non-Hermitian quantum systems.

quant-ph↗

arXiv2022

Anomalous low-energy $E2$-related behavior in triaxial nuclei

The anomalous low-energy $E2$-related behavior of a triaxially-deformed nucleus bas been identified and analyzed based on the SU(3) algebraic theory within the framework of the interacting boson model. The results show striking features that include a $B(E2;4_1^+\rightarrow2_1^+)/B(2_1^+\rightarrow0_1^+)<1.0$ transition rate ratio and a $E(4_1^+)/E(2_1^+)>2.0$ excitation energy ratio that can be tracked back to a finite-$N$ effect, which in a large-$N$ limit of the theory yields normal results for a stable $γ$-deformation. This description is shown to be able to explain observed $E2$ anomalous phenomenon in neutron-deficient nuclei such as $^{172}$Pt and $^{168}$Os, and in so doing yields a deeper understanding of the physical features of a soft triaxially-deformed nucleus.

nucl-th↗