Searcharxiv⌕ Search

arXiv subjects

Jing Shi

Publications and source records attributed to Jing Shi.

At least 91 records · Page 5Linked to original sources
arXiv2020

The 2020 ESPnet update: new features, broadened applications, performance improvements, and future plans

This paper describes the recent development of ESPnet (https://github.com/espnet/espnet), an end-to-end speech processing toolkit. This project was initiated in December 2017 to mainly deal with end-to-end speech recognition experiments based on sequence-to-sequence modeling. The project has grown rapidly and now covers a wide range of speech processing applications. Now ESPnet also includes text to speech (TTS), voice conversation (VC), speech translation (ST), and speech enhancement (SE) with support for beamforming, speech separation, denoising, and dereverberation. All applications are trained in an end-to-end manner, thanks to the generic sequence to sequence modeling properties, and they can be further integrated and jointly optimized. Also, ESPnet provides reproducible all-in-one recipes for these applications with state-of-the-art performance in various benchmarks by incorporating transformer, advanced data augmentation, and conformer. This project aims to provide up-to-date speech processing experience to the community so that researchers in academia and various industry scales can develop their technologies collaboratively.

eess.AS↗
arXiv2020

If Global or Local Investor Sentiments are Prone to Developing an Impact on Stock Returns, is there an Industry Effect?

This paper investigates the heterogeneous impacts of either Global or Local Investor Sentiments on stock returns. We study 10 industry sectors through the lens of 6 (so called) emerging countries: China, Brazil, India, Mexico, Indonesia and Turkey, over the 2000 to 2014 period. Using a panel data framework, our study sheds light on a significant effect of Local Investor Sentiments on expected returns for basic materials, consumer goods, industrial, and financial industries. Moreover, our results suggest that from Global Investor Sentiments alone, one cannot predict expected stock returns in these markets.

econ.GN↗
arXiv2020

ESPnet-se: end-to-end speech enhancement and separation toolkit designed for asr integration

We present ESPnet-SE, which is designed for the quick development of speech enhancement and speech separation systems in a single framework, along with the optional downstream speech recognition module. ESPnet-SE is a new project which integrates rich automatic speech recognition related models, resources and systems to support and validate the proposed front-end implementation (i.e. speech enhancement and separation).It is capable of processing both single-channel and multi-channel data, with various functionalities including dereverberation, denoising and source separation. We provide all-in-one recipes including data pre-processing, feature extraction, training and evaluation pipelines for a wide range of benchmark datasets. This paper describes the design of the toolkit, several important functionalities, especially the speech recognition integration, which differentiates ESPnet-SE from other open source toolkits, and experimental results with major benchmark datasets.

eess.AS↗
arXiv2020

Recent Developments on ESPnet Toolkit Boosted by Conformer

In this study, we present recent developments on ESPnet: End-to-End Speech Processing toolkit, which mainly involves a recently proposed architecture called Conformer, Convolution-augmented Transformer. This paper shows the results for a wide range of end-to-end speech processing applications, such as automatic speech recognition (ASR), speech translations (ST), speech separation (SS) and text-to-speech (TTS). Our experiments reveal various training tips and significant performance benefits obtained with the Conformer on different tasks. These results are competitive or even outperform the current state-of-art Transformer models. We are preparing to release all-in-one recipes using open source and publicly available corpora for all the above tasks with pre-trained models. Our aim for this work is to contribute to our research community by reducing the burden of preparing state-of-the-art research environments usually requiring high resources.

eess.AS↗
arXiv2020

Observation of magnon-polarons in a uniaxial antiferromagnetic insulator

Magnon-polarons, a type of hybridized excitations between magnons and phonons, were first reported in yttrium iron garnet as anomalies in the spin Seebeck effect responses. Here we report an observation of antiferromagnetic (AFM) magnon-polarons in a uniaxial AFM insulator Cr2O3. Despite the relatively higher energy of magnon than that of the acoustic phonons, near the spin-flop transition of ~ 6 T, the left-handed magnon spectrum shifts downward to hybridize with the acoustic phonons to form AFM magnon-polarons, which can also be probed by the spin Seebeck effect. The spin Seebeck signal is founded to be enhanced due to the magnon-polarons at low temperatures.

cond-mat.mtrl-sci↗
arXiv2020

A Benchmark and Baseline for Language-Driven Image Editing

Language-driven image editing can significantly save the laborious image editing work and be friendly to the photography novice. However, most similar work can only deal with a specific image domain or can only do global retouching. To solve this new task, we first present a new language-driven image editing dataset that supports both local and global editing with editing operation and mask annotations. Besides, we also propose a baseline method that fully utilizes the annotation to solve this problem. Our new method treats each editing operation as a sub-module and can automatically predict operation parameters. Not only performing well on challenging user data, but such an approach is also highly interpretable. We believe our work, including both the benchmark and the baseline, will advance the image editing area towards a more general and free-form level.

cs.CV↗
arXiv2020

Cubic Spline Smoothing Compensation for Irregularly Sampled Sequences

The marriage of recurrent neural networks and neural ordinary differential networks (ODE-RNN) is effective in modeling irregularly-observed sequences. While ODE produces the smooth hidden states between observation intervals, the RNN will trigger a hidden state jump when a new observation arrives, thus cause the interpolation discontinuity problem. To address this issue, we propose the cubic spline smoothing compensation, which is a stand-alone module upon either the output or the hidden state of ODE-RNN and can be trained end-to-end. We derive its analytical solution and provide its theoretical interpolation error bound. Extensive experiments indicate its merits over both ODE-RNN and cubic spline interpolation.

cs.LG↗
arXiv2020

Valley lifetimes of conduction band electrons in monolayer WSe$_2$

One of the main tasks in the investigation of 2-dimensional transition metal dichalcogenides is the determination of valley lifetimes. In this work, we combine time-resolved Kerr rotation with electrical transport measurements to explore the gate-dependent valley lifetimes of free conduction band electrons of monolayer WSe$_2$. When tuning the Fermi energy into the conduction band we observe a strong decrease of the respective valley lifetimes which is consistent with both spin-orbit and electron-phonon scattering. We explain the formation of a valley polarization by the scattering of optically excited valley polarized bright trions into dark states by intervalley scattering. Furthermore, we show that the conventional time-resolved Kerr rotation measurement scheme has to be modified to account for photo-induced gate screening effects. Disregarding this adaptation can lead to erroneous conclusions drawn from gate-dependent optical measurements and can completely mask the true gate-dependent valley dynamics.

cond-mat.mes-hall↗
arXiv2020

Speaker-Conditional Chain Model for Speech Separation and Extraction

Speech separation has been extensively explored to tackle the cocktail party problem. However, these studies are still far from having enough generalization capabilities for real scenarios. In this work, we raise a common strategy named Speaker-Conditional Chain Model to process complex speech recordings. In the proposed method, our model first infers the identities of variable numbers of speakers from the observation based on a sequence-to-sequence model. Then, it takes the information from the inferred speakers as conditions to extract their speech sources. With the predicted speaker information from whole observation, our model is helpful to solve the problem of conventional speech separation and speaker extraction for multi-round long recordings. The experiments from standard fully-overlapped speech separation benchmarks show comparable results with prior studies, while our proposed model gets better adaptability for multi-round long recordings.

eess.AS↗
arXiv2020

Sequence to Multi-Sequence Learning via Conditional Chain Mapping for Mixture Signals

Neural sequence-to-sequence models are well established for applications which can be cast as mapping a single input sequence into a single output sequence. In this work, we focus on one-to-many sequence transduction problems, such as extracting multiple sequential sources from a mixture sequence. We extend the standard sequence-to-sequence model to a conditional multi-sequence model, which explicitly models the relevance between multiple output sequences with the probabilistic chain rule. Based on this extension, our model can conditionally infer output sequences one-by-one by making use of both input and previously-estimated contextual output sequences. This model additionally has a simple and efficient stop criterion for the end of the transduction, making it able to infer the variable number of output sequences. We take speech data as a primary test field to evaluate our methods since the observed speech data is often composed of multiple sources due to the nature of the superposition principle of sound waves. Experiments on several different tasks including speech separation and multi-speaker speech recognition show that our conditional multi-sequence models lead to consistent improvements over the conventional non-conditional models.

eess.AS↗
arXiv2020

Current-induced CrI3 surface spin-flop transition probed by proximity magnetoresistance in Pt

By exploiting proximity coupling, we probe the spin state of the surface layers of CrI3, a van der Waals magnetic semiconductor, by measuring the induced magnetoresistance (MR) of Pt in Pt/CrI3 nano-devices. We fabricate the devices with clean and stable interfaces by placing freshly exfoliated CrI3 flake atop pre-patterned thin Pt strip and encapsulating the Pt/CrI3 heterostructure with hexagonal boron nitride (hBN) in a protected environment. In devices consisting of a wide range of CrI3 thicknesses (30 to 150 nm), we observe that an abrupt upward jump in Pt MR emerge at a 2 T magnetic field applied perpendicularly to the layers when the current density exceeds 2.5x10^10 A/m2, followed by a gradual decrease over a range of 5 T. These distinct MR features suggest a spin-flop transition which reveals strong antiferromagnetic interlayer coupling in the surface layers of CrI3. We study the current dependence by holding the Pt/CrI3 sample at approximately the same temperature to exclude the joule heating effect, and find that the MR jump increases with the current density, indicating a spin current origin. This spin current effect provides a new route to control spin configurations in insulating antiferromagnets, which is potentially useful for spintronic applications.

cond-mat.mes-hall↗
arXiv2020

Neural Speaker Diarization with Speaker-Wise Chain Rule

Speaker diarization is an essential step for processing multi-speaker audio. Although an end-to-end neural diarization (EEND) method achieved state-of-the-art performance, it is limited to a fixed number of speakers. In this paper, we solve this fixed number of speaker issue by a novel speaker-wise conditional inference method based on the probabilistic chain rule. In the proposed method, each speaker's speech activity is regarded as a single random variable, and is estimated sequentially conditioned on previously estimated other speakers' speech activities. Similar to other sequence-to-sequence models, the proposed method produces a variable number of speakers with a stop sequence condition. We evaluated the proposed method on multi-speaker audio recordings of a variable number of speakers. Experimental results show that the proposed method can correctly produce diarization results with a variable number of speakers and outperforms the state-of-the-art end-to-end speaker diarization methods in terms of diarization error rate.

eess.AS↗
arXiv2020

Spin Seebeck Effect near the Antiferromagnetic Spin-Flop Transition

We develop a low-temperature, long-wavelength theory for the interfacial spin Seebeck effect (SSE) in easy-axis antiferromagnets. The field-induced spin-flop (SF) transition of Néel order is associated with a qualitative change in SSE behavior: Below SF, there are two spin carriers with opposite magnetic moments, with the carriers polarized along the field forming a majority magnon band. Above SF, the low-energy, ferromagnetic-like mode has magnetic moment opposite the field. This results in a sign change of the SSE across SF, which agrees with recent measurements on Cr$_2$O$_3$/Pt and Cr$_2$O$_3$/Ta devices [Li $\textit{et al.,}$ $\textit{Nature}$ $\textbf{578,}$ 70 (2020)]. In our theory, SSE is due to a Néel spin current below SF and a magnetic spin current above SF. Using the ratio of the associated Néel to magnetic spin-mixing conductances as a single constant fitting parameter, we reproduce the field dependence of the experimental data and partially the temperature dependence of the relative SSE jump across SF.

cond-mat.mes-hall↗
arXiv2020

Unveiling valley lifetimes of free charge carriers in monolayer WSe$_2$

We report on nanosecond long, gate-dependent valley lifetimes of free charge carriers in monolayer WSe$_2$, unambiguously identified by the combination of time-resolved Kerr rotation and electrical transport measurements. While the valley polarization increases when tuning the Fermi level into the conduction or valence band, there is a strong decrease of the respective valley lifetime consistent with both electron-phonon and spin-orbit scattering. The longest lifetimes are seen for spin-polarized bound excitons in the band gap region. We explain our findings via two distinct, Fermi level-dependent scattering channels of optically excited, valley polarized bright trions either via dark or bound states. By electrostatic gating we demonstrate that the transition metal dichalcogenide WSe$_2$ can be tuned to be either an ideal host for long-lived localized spin states or allow for nanosecond valley lifetimes of free charge carriers (> 10 ns).

cond-mat.mes-hall↗
arXiv2019

Coexistence of Magnetic Orders in Two-Dimensional Magnet CrI$_3$

The magnetic properties in two-dimensional van der Waals materials depend sensitively on structure. CrI3, as an example, has been recently demonstrated to exhibit distinct magnetic properties depending on the layer thickness and stacking order. Bulk CrI3 is ferromagnetic (FM) with a Curie temperature of 61 K and a rhombohedral layer stacking, while few-layer CrI3 has a layered antiferromagnetic (AFM) phase with a lower ordering temperature of 45 K and a monoclinic stacking. In this work, we use cryogenic magnetic force microscopy to investigate CrI3 flakes in the intermediate thickness range (25 - 200 nm) and find that the two types of magnetic orders hence the stacking orders can coexist in the same flake, with a layer of ~13 nm at each surface being in the layered AFM phase similar to few-layer CrI3 and the rest in the bulk FM phase. The switching of the bulk moment proceeds through a remnant state with nearly compensated magnetic moment along the c-axis, indicating formation of c-axis domains allowed by a weak interlayer coupling strength in the rhombohedral phase. Our results provide a comprehensive picture on the magnetism in CrI3 and point to the possibility of engineering magnetic heterostructures within the same material.

cond-mat.mes-hall↗
arXiv2019

Fe$_{5-x}$Ge$_2$Te$_2$ -- A new exfoliable itinerant ferromagnet with high Curie temperature and large perpendicular magnetic anisotropy

Layered van der Waals (vdW) crystals with intrinsic magnetic properties such as high Curie temperature (TC) and large perpendicular magnetic anisotropy (PMA) are key to the development and application of spintronic devices. The ferromagnetic vdW metal Fe3-xGeTe2 (FGT) has gained prominence recently due to its high TC (220 K) and strong PMA. Here, we introduce a new metallic vdW ferromagnets, Fe5-xGe2Te2 or FG2T, which was successfully synthesized and fully characterized. FG2T is a metal that orders ferromagnetically with a very sharp transition at 250 K (bulk and single crystal thin flakes) and shows large PMA, as found by both experimental and computational studies. This work enables novel heterostructure devices with near room temperature capabilities by using FG2T as spin injector.

cond-mat.mtrl-sci↗
arXiv2019

Spin Seebeck effect from antiferromagnetic magnons and critical spin fluctuations in epitaxial FeF2 films

We report a longitudinal spin Seebeck effect (SSE) study in epitaxially grown FeF2(110) antiferromagnetic (AFM) thin films with strong uniaxial anisotropy over the temperature range of 3.8 - 250 K. Both the magnetic field- and temperature-dependent SSE signals below the Néel temperature (TN=70 K) of the FeF2 films are consistent with a theoretical model based on the excitations of AFM magnons without any net induced static magnetic moment. In addition to the characteristic low-temperature SSE peak associated with the AFM magnons, there is another SSE peak at TN which extends well into the paramagnetic phase. All the SSE data taken at different magnetic fields up to 12 T near and above the critical point TN follow the critical scaling law very well with the critical exponents for magnetic susceptibility of 3D Ising systems, which suggests that the AFM spin correlation is responsible for the observed SSE near TN.

cond-mat.mtrl-sci↗