arXiv subjects

Jinyu Liu

Publications and source records attributed to Jinyu Liu.

At least 37 records · Page 2Linked to original sources

arXiv2023

Two-Stream Joint-Training for Speaker Independent Acoustic-to-Articulatory Inversion

Acoustic-to-articulatory inversion (AAI) aims to estimate the parameters of articulators from speech audio. There are two common challenges in AAI, which are the limited data and the unsatisfactory performance in speaker independent scenario. Most current works focus on extracting features directly from speech and ignoring the importance of phoneme information which may limit the performance of AAI. To this end, we propose a novel network called SPN that uses two different streams to carry out the AAI task. Firstly, to improve the performance of speaker-independent experiment, we propose a new phoneme stream network to estimate the articulatory parameters as the phoneme features. To the best of our knowledge, this is the first work that extracts the speaker-independent features from phonemes to improve the performance of AAI. Secondly, in order to better represent the speech information, we train a speech stream network to combine the local features and the global features. Compared with state-of-the-art (SOTA), the proposed method reduces 0.18mm on RMSE and increases 6.0% on Pearson correlation coefficient in the speaker-independent experiment. The code has been released at https://github.com/liujinyu123/AAINetwork-SPN.

cs.SD↗

arXiv2023

Magnetic tuning of band topology evidenced by exotic quantum oscillations in the Dirac semimetal EuMnSb$_2$

Interplay between magnetism and electronic band topology is of central current interest in topological matter research. We use quantum oscillations as powerful tool to probe the evolution of band topology in the Dirac semimetal EuMnSb$_2$. The Eu local 4f magnetic moments display different antiferromagnetic states below 25 K and a field-polarized phase above 16 T. Upon cooling from 65 K into the field-polarized state, an exotic temperature dependent shift of oscillation peaks arises, accompanied by the development of non-zero Berry phase and a huge unconventional splitting of the oscillations. Band-structure calculations confirm the change from trivial to non-trivial band topology induced by the ferromagnetic Eu state, classifying EuMnSb$_2$ as unique magnetic topological semimetal.

cond-mat.str-el↗

arXiv2022

Modality-Aware Contrastive Instance Learning with Self-Distillation for Weakly-Supervised Audio-Visual Violence Detection

Weakly-supervised audio-visual violence detection aims to distinguish snippets containing multimodal violence events with video-level labels. Many prior works perform audio-visual integration and interaction in an early or intermediate manner, yet overlooking the modality heterogeneousness over the weakly-supervised setting. In this paper, we analyze the modality asynchrony and undifferentiated instances phenomena of the multiple instance learning (MIL) procedure, and further investigate its negative impact on weakly-supervised audio-visual learning. To address these issues, we propose a modality-aware contrastive instance learning with self-distillation (MACIL-SD) strategy. Specifically, we leverage a lightweight two-stream network to generate audio and visual bags, in which unimodal background, violent, and normal instances are clustered into semi-bags in an unsupervised way. Then audio and visual violent semi-bag representations are assembled as positive pairs, and violent semi-bags are combined with background and normal instances in the opposite modality as contrastive negative pairs. Furthermore, a self-distillation module is applied to transfer unimodal visual knowledge to the audio-visual model, which alleviates noises and closes the semantic gap between unimodal and multimodal features. Experiments show that our framework outperforms previous methods with lower complexity on the large-scale XD-Violence dataset. Results also demonstrate that our proposed approach can be used as plug-in modules to enhance other networks. Codes are available at https://github.com/JustinYuu/MACIL_SD.

cs.CV↗

arXiv2022

Self-Supervised Video Representation Learning with Motion-Contrastive Perception

Visual-only self-supervised learning has achieved significant improvement in video representation learning. Existing related methods encourage models to learn video representations by utilizing contrastive learning or designing specific pretext tasks. However, some models are likely to focus on the background, which is unimportant for learning video representations. To alleviate this problem, we propose a new view called long-range residual frame to obtain more motion-specific information. Based on this, we propose the Motion-Contrastive Perception Network (MCPNet), which consists of two branches, namely, Motion Information Perception (MIP) and Contrastive Instance Perception (CIP), to learn generic video representations by focusing on the changing areas in videos. Specifically, the MIP branch aims to learn fine-grained motion features, and the CIP branch performs contrastive learning to learn overall semantics information for each instance. Experiments on two benchmark datasets UCF-101 and HMDB-51 show that our method outperforms current state-of-the-art visual-only self-supervised approaches.

cs.CV↗

arXiv2022

Acoustic-to-articulatory Inversion based on Speech Decomposition and Auxiliary Feature

Acoustic-to-articulatory inversion (AAI) is to obtain the movement of articulators from speech signals. Until now, achieving a speaker-independent AAI remains a challenge given the limited data. Besides, most current works only use audio speech as input, causing an inevitable performance bottleneck. To solve these problems, firstly, we pre-train a speech decomposition network to decompose audio speech into speaker embedding and content embedding as the new personalized speech features to adapt to the speaker-independent case. Secondly, to further improve the AAI, we propose a novel auxiliary feature network to estimate the lip auxiliary features from the above personalized speech features. Experimental results on three public datasets show that, compared with the state-of-the-art only using the audio speech feature, the proposed method reduces the average RMSE by 0.25 and increases the average correlation coefficient by 2.0% in the speaker-dependent case. More importantly, the average RMSE decreases by 0.29 and the average correlation coefficient increases by 5.0% in the speaker-independent case.

cs.SD↗

arXiv2021

Leveraging Large-scale Computational Database and Deep Learning for Accurate Prediction of Material Properties

Accurately predicting the physical and chemical properties of materials remains one of the most challenging tasks in material design, and one effective strategy is to construct a reliable data set and use it for training a machine learning model. In this study, we constructed a large-scale material genome database (Matgen) containing 76,463 materials collected from experimentally-observed database, and computed their bandgap properties through the Density functional theory (DFT) method with Perdew-Burke-Ernzehof (PBE) functional. We verified the computation method by comparing part of our results with those from the open Material Project (MP) and Open Quantum Materials Database (OQMD), all with PBE computations, and found that Matgen achieved the same computation accuracy based on both measured and computed bandgap properties. Based on the computed properties of our comprehensive dataset, we have developed a new graph-based deep learning model, namely CrystalNet, through our recently developed Communicative Message Passing Neural Network (CMPNN) framework. The model was shown to outperform other state-of-the-art prediction models. A further fine-tuning on 1716 experimental bandgap values (CrystalNet-TL) achieved a superior performance with mean absolute error (MAE) of 0.77 eV on independent test, which has outperformed the pure PBE (1.14~1.45 eV). Moreover, the model was proven applicable to hypothetical materials with MAE of 0.77 eV as referred by computations from HSE, a highly accurate quantum mechanics (QM) method, consist better than PBE (MAE=1.13eV). We also made material structures, computed properties by PBE, and the CrystalNet models publically available at https://matgen.nscc-gz.cn.

cond-mat.mtrl-sci↗

arXiv2021

Bid Optimization using Maximum Entropy Reinforcement Learning

Real-time bidding (RTB) has become a critical way of online advertising. In RTB, an advertiser can participate in bidding ad impressions to display its advertisements. The advertiser determines every impression's bidding price according to its bidding strategy. Therefore, a good bidding strategy can help advertisers improve cost efficiency. This paper focuses on optimizing a single advertiser's bidding strategy using reinforcement learning (RL) in RTB. Unfortunately, it is challenging to optimize the bidding strategy through RL at the granularity of impression due to the highly dynamic nature of the RTB environment. In this paper, we first utilize a widely accepted linear bidding function to compute every impression's base price and optimize it by a mutable adjustment factor derived from the RTB auction environment, to avoid optimizing every impression's bidding price directly. Specifically, we use the maximum entropy RL algorithm (Soft Actor-Critic) to optimize the adjustment factor generation policy at the impression-grained level. Finally, the empirical study on a public dataset demonstrates that the proposed bidding strategy has superior performance compared with the baselines.

cs.LG↗

arXiv2020

Tuning quantum transport by controlling spin reorientations in Dirac semimetal candidates Eu$_{1-x}$Sr$_{x}$MnSb$_{2}$

Magnetic topological semimetals have attracted intense attention recently since these materials carry a great promise for potential applications in novel spintronic devices. Here, we report an intimate interplay between lattice, Eu magnetic order and topological semimetallic behavior in Eu$_{1-x}$Sr$_{x}$MnSb$_{2}$ driven by nonmagnetic Sr doping on magnetic Eu site. Different types of Eu spin reorientations are controllable by the Sr concentration, temperature or magnetic field, and coupled to the quantum transport properties of Dirac fermions generated by the 2D Sb layers. Our study opens a new pathway to achieving exotic magnetic order and topological semimetallic state via controlling spin reorientation. The effective strategy of substituting rare-earth site by nonmagnetic element demonstrated here may be applicable to the AMnCh$_{2}$ (A=rare-earth elements; Ch=Bi/Sb) family and a wide variation of other layered compounds involving spatially separated rare-earth and transition metal layers.

cond-mat.str-el↗

arXiv2020

Quantum oscillation and unusual protection mechanism of the surface state in nonsymmorphic semimetals

In a topological semimetal with Dirac or Weyl points, the bulk edge correspondence principle predicts a gapless edge mode if the essential symmetry is still preserved at the surface. The detection of such topological surface state has been considered as the fingerprint prove for crystals with nontrivial topological bulk band. On the contrary, it has been proposed that even with symmetry broken at the surface, a new surface band can emerge in nonsymmorphic topological semimetals. The symmetry reduction at the surface lifts the bulk band degeneracies, produces an unusual floating surface band with trivial topology. Here, we report quantum transport probing to ZrSiSe thin flakes and reveal transport signatures of this new surface state. Remarkably, though topologically trivial, such a surface band exhibit substantial two dimensional Shubnikov de Haas quantum oscillations with high mobility, which signifies a new protection mechanism and may open applications for surface-related devices.

cond-mat.mtrl-sci↗

arXiv2020

A Néel-type antiferromagnetic order in the spin 1/2 rare-earth honeycomb YbCl$_3$

Most of the searches for Kitaev materials deal with $4d/5d$ magnets with spin-orbit-coupled ${J=1/2}$ local moments such as iridates and $α$-RuCl$_3$. Here we propose the monoclinic YbCl$_3$ with a Yb$^{3+}$ honeycomb lattice for the exploration of Kiteav physics. We perform thermodynamic, $ac$ susceptibility, angle-dependent magnetic torque and neutron diffraction measurements on YbCl$_3$ single crystal. We find that the Yb$^{3+}$ ion exhibits a Kramers doublet ground state that gives rise to an effective spin ${J_{\text{eff}}=1/2}$ local moment. The compound exhibits short-range magnetic order below 1.20 K, followed by a long-range Néel-type antiferromagnetic order at 0.60 K, below which the ordered Yb$^{3+}$ spins lie in the $ac$ plane with an angle of 16(11)$^{\circ}$ away from the $a$ axis. These orders can be suppressed by in-plane and out-of-plane magnetic fields at around 6 and 10 T, respectively. Moreover, the Néel temperature varies non-monotonically under the out-of-plane magnetic fields. The in-plane magnetic anisotropy and the reduced order moment 0.8(1) $μ_B$ at 0.25 K indicate that YbCl$_3$ could be a two-dimensional spin system to proximate the Kitaev physics.

cond-mat.str-el↗

arXiv2019

High yield production of ultrathin fibroid semiconducting nanowire of Ta$_2$Pd$_3$Se$_8$

Immediately after the demonstration of the high-quality electronic properties in various two dimensional (2D) van der Waals (vdW) crystals fabricated with mechanical exfoliation, many methods have been reported to explore and control large scale fabrications. Comparing with recent advancements in fabricating 2D atomic layered crystals, large scale production of one dimensional (1D) nanowires with thickness approaching molecular or atomic level still remains stagnant. Here, we demonstrate the high yield production of a 1D vdW material, semiconducting Ta2Pd3Se8 nanowires, by means of liquid-phase exfoliation. The thinnest nanowire we have readily achieved is around 1 nm, corresponding to a bundle of one or two molecular ribbons. Transmission electron microscopy and transport measurements reveal the as-fabricated Ta2Pd3Se8 nanowires exhibit unexpected high crystallinity and chemical stability. Our low frequency Raman spectroscopy reveals clear evidence of the existing of weak inter-ribbon bindings. The fabricated nanowire transistors exhibit high switching performance and promising applications for photodetectors.

physics.app-ph↗

arXiv2019

A van der Waals antiferromagnetic topological insulator with weak interlayer magnetic coupling

Magnetic topological insulators (TI) provide an important material platform to explore quantum phenomena such as quantized anomalous Hall (QAH) effect and Majorana modes, etc. Their successful material realization is thus essential for our fundamental understanding and potential technical revolutions. By realizing a bulk van der Waals material MnBi4Te7 with alternating septuple [MnBi2Te4] and quintuple [Bi2Te3] layers, we show that it is ferromagnetic in plane but antiferromagnetic along the c axis with an out-of-plane saturation field of ~ 0.22 T at 2 K. Our angle-resolved photoemission spectroscopy measurements and first-principles calculations further demonstrate that MnBi4Te7 is a Z2 antiferromagnetic TI with two types of surface states associated with the [MnBi2Te4] or [Bi2Te3] termination, respectively. Additionally, its superlattice nature may make various heterostructures of [MnBi2Te4] and [Bi2Te3] layers possible by exfoliation. Therefore, the low saturation field and the superlattice nature of MnBi4Te7 make it an ideal system to investigate rich emergent phenomena.

cond-mat.mtrl-sci↗

arXiv2019

Non-trivial topology in a layered Dirac nodal-line semimetal candidate SrZnSb$_2$ with distorted Sb square nets

Dirac states hosted by Sb/Bi square nets are known to exist in the layered antiferromagnetic AMnX$_2$ (A = Ca/Sr/Ba/Eu/Yb, X=Sb/Bi) material family the space group to be P4/nmm or I4/mmm. In this paper, we present a comprehensive study of quantum transport behaviors, angle-resolved photoemission spectroscopy (ARPES) and first-principles calculations on SrZnSb2, a nonmagnetic analogue to AMnX2, which crystallizes in the pnma space group with distorted square nets. From the quantum oscillation measurements up to 35 T, three major frequencies including F$_1$ = 103 T, F$_2$ = 127 T and F$_3$ = 160 T, are identified. The effective masses of the quasiparticles associated with these frequencies are extracted, namely, m*$_1$ = 0.1 m$_e$, m*$_2$ = 0.1 m$_e$ and m*$_3$ = 0.09m$_e$, where m$_e$ is the free electron mass. From the three-band Lifshitz-Kosevich fit, the Berry phases accumulated along the cyclotron orbit of the quasiparticles are 0.06$π$, 1.2$π$ and 0.74$π$ for F$_1$, F$_2$ and F$_3$, respectively. Combined with the ARPES data and the first-principles calculations, we reveal that F2 and F3 are associated with the two nontrivial Fermi pockets at the Brillouin zone edge while F1 is associated with the trivial Fermi pocket at the zone center. In addition, the first-principles calculations further suggest the existence of Dirac nodal line in the band structure of SrZnSb$_2$.

cond-mat.mtrl-sci↗

arXiv2019

Quasi-2D magnetism and origin of the Dirac semimetallic behavior in nonstoichiometric Sr$_{1-y}$Mn$_{1-z}$Sb$_{2}$ (y, z~$<$0.1)

Nonstoichiometric Sr$_{1-y}$Mn$_{1-z}$Sb$_{2}$ (y, z~$<$0.1) is known to exhibit a coexistence of magnetic order and the nontrivial semimetallic behavior related to Dirac or Weyl fermions. Here, we report inelastic neutron scattering analyses of the spin dynamics and density functional theory studies on the electronic properties of Sr$_{1-y}$Mn$_{1-z}$Sb$_{2}$. We observe a relatively large spin excitation gap $\approx$ 8.5 meV at 5 K, and the interlayer magnetic exchange constant only 2.8 \% of the dominant intralayer magnetic interaction, providing evidence that Sr$_{1-y}$Mn$_{1-z}$Sb$_{2}$ exhibits a quasi-2D magnetism. Using density functional theory, we find a strong influence of magnetic orders on the electronic band structure and the Dirac dispersions near the Fermi level along the Y-S direction in the presence of a ferromagnetic ordering. Our study unveils novel interplay between the magnetic order, magnetic transition, and electronic property in Sr$_{1-y}$Mn$_{1-z}$Sb$_{2}$, and opens new pathways to control the relativistic band structure through magnetism in ternary compounds.

cond-mat.str-el↗

arXiv2018

Lithium ion intercalation in thin crystals of hexagonal TaSe2 gated by a polymer electrolyte

Ionic liquid gating has been used to modify properties of layered transition metal dichalcogenides (TMDCs), including two-dimensional (2D) crystals of TMDCs used extensively recently in the device work, which has led to observations of properties not seen in the bulk. The main effect comes from the electrostatic gating due to strong electric field at the interface. In addition, ionic liquid gating also leads to ion intercalation when the ion size of gate electrolyte is small compared to the interlayer spacing of TMDCs. However, the microscopic processes of ion intercalation have rarely been explored in layered TMDCs. Here, we employed a technique combining photolithography device fabrication and electrical transport measurements on the thin crystals of hexagonal TaSe2 using multiple channel devices gated by a polymer electrolyte LiClO4/PEO. The gate voltage and time dependent source-drain resistances of these thin crystals were used to obtain information on the intercalation process, the effect of ion intercalation, and the correlation between the ion occupation of allowed interstitial sites and the device characteristics. We found a gate voltage controlled modulation of the charge density waves and scattering rate of charge carriers. Our work suggests that ion intercalation can be a useful tool for layered materials engineering and 2D crystal device design.

cond-mat.mtrl-sci↗

arXiv2018

Angle-dependent magnetoresistance as a sensitive probe of the charge density wave in quasi-one-dimensional semimetal Ta$_2$NiSe$_7$

The behavior of charge density wave (CDW) in an external magnetic field is dictated by both orbital and Pauli (Zeeman) effects. A quasi-one-dimensional (Q1D) system features Q1D Fermi surfaces that allow these effects to be distinguished, which in turn can provide sensitive probe to the underlying electronic states. Here we studied the field dependence of an incommensurate CDW in a transition-metal chalcogenide Ta2NiSe7 with a Q1D chain structure. The angle-dependent magnetoresistance (MR) is found to be very sensitive to the relative orientation between the magnetic field and the chain direction. With an applied current fixed along the b axis (the chain direction), the angle-dependent MR shows a striking change of the symmetry below T_CDW only for a rotating magnetic field in the ac plane. In contrast, the symmetry axis remains unchanged for other configurations (H in ab and bc plane). The orbital effect conforms to the lattice symmetry, while Pauli effect in the form of μB B / v_F can be responsible for such symmetry change, provided that the Fermi velocity v_F is significantly anisotropic and the nesting vector changes in a magnetic field, which is corroborated by our first-principles calculations. Our results show that the angle-dependent MR is a sensitive transport probe of CDW and can be useful for the study of low-dimensional systems in general.

cond-mat.str-el↗

arXiv2018

Spin-flop phase transition in the orthorhombic antiferromagnetic topological semimetal Cu0.95MnAs

The orthorhombic antiferromagnetic compound CuMnAs was recently predicted to be an antiferromagnetic Dirac semimetal if both the Ry gliding and S2z rotational symmetries are preserved in its magnetic ordered state. In our previous work on Cu0.95MnAs and Cu0.98Mn0.96As, we showed that in their low temperature commensurate antiferromagnetic state the b axis is the magnetic easy axis, which breaks the S2z symmetry. As a result, while the existence of Dirac fermions is no longer protected, the polarized surface state makes this material potentially interesting for antiferromagnetic spintronics. In this paper, we report a detailed study of the anisotropic magnetic properties and magnetoresistance of Cu0.95MnAs and Cu0.98Mn0.96As. Our study shows that in Cu0.95MnAs the b axis is the easy axis and the c axis is the hard axis. Furthermore, it reveals that Cu0.95MnAs features a spin-flop phase transition at high temperatures and low fields when the field is applied along the easy b axis, resulting in canted antiferromagnetism. However, no metamagnetic transition is observed in Cu0.98Mn0.96As, indicating that the magnetic interactions in this system are very sensitive to Cu vacancies and Cu/Mn site mixing.

cond-mat.str-el↗

arXiv2018

Band dependence of charge density wave in quasi-one-dimensional Ta2NiSe7 probed by orbital magnetoresistance

Ta2NiSe7 is a quasi-one-dimensional (quasi-1D) transition-metal chalcogenide with Ta and Ni chain structure. An incommensurate charge-density wave (CDW) in this quasi-1D structure was well studied previously using tunnelling spectrum, X-ray and electron diffraction, whereas its transport property and the relation to the underlying electronic states remain to be explored. Here we report our results of magnetoresistance (MR) on Ta2NiSe7. A breakdown of the Kohler's rule is found upon entering the CDW state. Concomitantly, a clear change of curvature in the field dependence of MR is observed. We show that the curvature change is well described by two-band orbital MR, with the hole density being strongly suppressed in the CDW state, indicating that the $p$ orbitals from Se atoms dominate the change in transport through the CDW transition.

cond-mat.mtrl-sci↗