arXiv subjects

Yuxuan Wang

Publications and source records attributed to Yuxuan Wang.

At least 289 records · Page 16Linked to original sources

arXiv2021

VeniBot: Towards Autonomous Venipuncture with Automatic Puncture Area and Angle Regression from NIR Images

Venipucture is a common step in clinical scenarios, and is with highly practical value to be automated with robotics. Nowadays, only a few on-shelf robotic systems are developed, however, they can not fulfill practical usage due to varied reasons. In this paper, we develop a compact venipucture robot -- VeniBot, with four parts, six motors and two imaging devices. For the automation, we focus on the positioning part and propose a Dual-In-Dual-Out network based on two-step learning and two-task learning, which can achieve fully automatic regression of the suitable puncture area and angle from near-infrared(NIR) images. The regressed suitable puncture area and angle can further navigate the positioning part of VeniBot, which is an important step towards a fully autonomous venipucture robot. Validation on 30 VeniBot-collected volunteers shows a high mean dice coefficient(DSC) of 0.7634 and a low angle error of 15.58° on suitable puncture area and angle regression respectively, indicating its potentially wide and practical application in the future.

eess.IV↗

arXiv2021

Phase diagram of the spin-1/2 Yukawa-SYK model: Non-Fermi liquid, insulator, and superconductor

We analyze the phase diagram of the Yukawa-Sachdev-Ye-Kitaev model, which describes complex fermions randomly interacting with real bosons via a Yukawa coupling, at finite temperatures and varying fermion density. In a recent work [Phys. Rev. Research 2, 033084 (2020)] it has been shown that, upon varying filling or chemical potential, there exists a first-order quantum phase transition between a non-Fermi liquid (nFL) phase and an insulating phase. Here we show that in such a model with time-reversal symmetry this quantum phase transition is preempted by a pairing phase that develops as a low-temperature instability. As a remnant of the would-be nFL-insulator transition, the superconducting critical temperature rapidly decreases beyond a certain chemical potential. On the other hand, depending on parameters the first-order quantum phase transition extend to finite-temperatures and terminate at a thermal critical point, beyond which the nFL and the insulator become the same phase, similar to that of the liquid-gas and metal-insulator transition in real materials. We determine the pairing phase boundary and the location of the thermal critical point via combined analytic and quantum Monte Carlo numeric efforts. Our results provide the model realization of the transition of nFL's towards superconductivity and insulating states, therefore offer a controlled platform for future investigations of the generic phase diagram that hosts nFL, insulator and superconductor and their phase transitions.

cond-mat.str-el↗

arXiv2021

Supervised Chorus Detection for Popular Music Using Convolutional Neural Network and Multi-task Learning

This paper presents a novel supervised approach to detecting the chorus segments in popular music. Traditional approaches to this task are mostly unsupervised, with pipelines designed to target some quality that is assumed to define "chorusness," which usually means seeking the loudest or most frequently repeated sections. We propose to use a convolutional neural network with a multi-task learning objective, which simultaneously fits two temporal activation curves: one indicating "chorusness" as a function of time, and the other the location of the boundaries. We also propose a post-processing method that jointly takes into account the chorus and boundary predictions to produce binary output. In experiments using three datasets, we compare our system to a set of public implementations of other segmentation and chorus-detection algorithms, and find our approach performs significantly better.

eess.AS↗

arXiv2021

Rebound and scattering of motile Chlamydomonas algae in confined chambers

Motivated by recent experiments demonstrating that motile algae get trapped in draining foams, we study the trajectories of microorganisms confined in model foam channels (section of a Plateau border). We track single Chlamydomonas reinhardtii cells confined in a thin three-circle microfluidic chamber and show that their spatial distribution exhibits strong corner accumulation. Using empirical scattering laws observed in previous experiments (scattering with a constant scattering angle), we next develop a two-dimension geometrical model and compute the phase space of trapped and periodic trajectories of swimmers inside a three-circles billiard. We find that the majority of cell trajectories end up in a corner, providing a geometrical mechanism for corner accumulation. Incorporating the distribution of scattering angles observed in our experiments and including hydrodynamic interactions between the cells and the surfaces into the geometrical model enables us to reproduce the experimental probability density function of micro-swimmers in microfluidic chambers. Both our experiments and models demonstrate therefore that motility leads generically to trapping in complex geometries.

physics.bio-ph↗

arXiv2021

Modeling the Compatibility of Stem Tracks to Generate Music Mashups

A music mashup combines audio elements from two or more songs to create a new work. To reduce the time and effort required to make them, researchers have developed algorithms that predict the compatibility of audio elements. Prior work has focused on mixing unaltered excerpts, but advances in source separation enable the creation of mashups from isolated stems (e.g., vocals, drums, bass, etc.). In this work, we take advantage of separated stems not just for creating mashups, but for training a model that predicts the mutual compatibility of groups of excerpts, using self-supervised and semi-supervised methods. Specifically, we first produce a random mashup creation pipeline that combines stem tracks obtained via source separation, with key and tempo automatically adjusted to match, since these are prerequisites for high-quality mashups. To train a model to predict compatibility, we use stem tracks obtained from the same song as positive examples, and random combinations of stems with key and/or tempo unadjusted as negative examples. To improve the model and use more data, we also train on "average" examples: random combinations with matching key and tempo, where we treat them as unlabeled data as their true compatibility is unknown. To determine whether the combined signal or the set of stem signals is more indicative of the quality of the result, we experiment on two model architectures and train them using semi-supervised learning technique. Finally, we conduct objective and subjective evaluations of the system, comparing them to a standard rule-based system.

cs.SD↗

arXiv2021

USTC-NELSLIP System Description for DIHARD-III Challenge

This system description describes our submission system to the Third DIHARD Speech Diarization Challenge. Besides the traditional clustering based system, the innovation of our system lies in the combination of various front-end techniques to solve the diarization problem, including speech separation and target-speaker based voice activity detection (TS-VAD), combined with iterative data purification. We also adopted audio domain classification to design domain-dependent processing. Finally, we performed post processing to do system fusion and selection. Our best system achieved DERs of 11.30% in track 1 and 16.78% in track 2 on evaluation set, respectively.

cs.SD↗

arXiv2021

Higher-order topological superconductors from Weyl semimetals

We propose that doped Weyl semimetals with four Weyl points are natural candidates to realize higher-order topological superconductors, which exhibit a fully gapped bulk while the surface hosts robust gapless chiral hinge states. We show that in such a doped Weyl semimetal, a featureless finite-range attractive interaction favors a $p+ip$ pairing symmetry. By analyzing its topological properties, we identify such a chiral pairing state as a higher-order topological superconductor, which depending on the existence of a four-fold roto-inversion symmetry $\mathsf{R}_{4z}$, is either intrinsic (meaning that the corresponding hinge states can only be removed by closing the bulk gap, rather than modifying the surface states) or extrinsic. We achieve this understanding via various methods recently developed for higher-order topology, including Wannier representability, Wannier spectrum, and defect classification approaches. For the $\mathsf{R}_{4z}$ symmetric case, we provide a complete classification of the higher-order topological superconductors. We show that such second-order topological superconductors exhibit chiral hinge modes that are robust in the absence of interaction effects but can be eliminated at the cost of introducing surface topological order.

cond-mat.supr-con↗

arXiv2021

CatNet: music source separation system with mix-audio augmentation

Music source separation (MSS) is the task of separating a music piece into individual sources, such as vocals and accompaniment. Recently, neural network based methods have been applied to address the MSS problem, and can be categorized into spectrogram and time-domain based methods. However, there is a lack of research of using complementary information of spectrogram and time-domain inputs for MSS. In this article, we propose a CatNet framework that concatenates a UNet separation branch using spectrogram as input and a WavUNet separation branch using time-domain waveform as input for MSS. We propose an end-to-end and fully differentiable system that incorporate spectrogram calculation into CatNet. In addition, we propose a novel mix-audio data augmentation method that randomly mix audio segments from the same source as augmented audio segments for training. Our proposed CatNet MSS system achieves a state-of-the-art vocals separation source distortion ratio (SDR) of 7.54 dB, outperforming MMDenseNet of 6.57 dB evaluated on the MUSDB18 dataset.

cs.SD↗

arXiv2021

Speech enhancement with weakly labelled data from AudioSet

Speech enhancement is a task to improve the intelligibility and perceptual quality of degraded speech signal. Recently, neural networks based methods have been applied to speech enhancement. However, many neural network based methods require noisy and clean speech pairs for training. We propose a speech enhancement framework that can be trained with large-scale weakly labelled AudioSet dataset. Weakly labelled data only contain audio tags of audio clips, but not the onset or offset times of speech. We first apply pretrained audio neural networks (PANNs) to detect anchor segments that contain speech or sound events in audio clips. Then, we randomly mix two detected anchor segments containing speech and sound events as a mixture, and build a conditional source separation network using PANNs predictions as soft conditions for speech enhancement. In inference, we input a noisy speech signal with the one-hot encoding of "Speech" as a condition to the trained system to predict enhanced speech. Our system achieves a PESQ of 2.28 and an SSNR of 8.75 dB on the VoiceBank-DEMAND dataset, outperforming the previous SEGAN system of 2.16 and 7.73 dB respectively.

cs.SD↗

arXiv2021

ByteSing: A Chinese Singing Voice Synthesis System Using Duration Allocated Encoder-Decoder Acoustic Models and WaveRNN Vocoders

This paper presents ByteSing, a Chinese singing voice synthesis (SVS) system based on duration allocated Tacotron-like acoustic models and WaveRNN neural vocoders. Different from the conventional SVS models, the proposed ByteSing employs Tacotron-like encoder-decoder structures as the acoustic models, in which the CBHG models and recurrent neural networks (RNNs) are explored as encoders and decoders respectively. Meanwhile an auxiliary phoneme duration prediction model is utilized to expand the input sequence, which can enhance the model controllable capacity, model stability and tempo prediction accuracy. WaveRNN neural vocoders are also adopted as neural vocoders to further improve the voice quality of synthesized songs. Both objective and subjective experimental results prove that the SVS method proposed in this paper can produce quite natural, expressive and high-fidelity songs by improving the pitch and spectrogram prediction accuracy and the models using attention mechanism can achieve best performance.

eess.AS↗

arXiv2021

Topological and nematic superconductivity mediated by ferro-SU(4) fluctuations in twisted bilayer graphene

We propose an SU(4) spin-valley-fermion model to investigate the superconducting instabilities of twisted bilayer graphene (TBG). In this approach, bosonic fluctuations associated with an emergent SU(4) symmetry, corresponding to combined rotations in valley and spin spaces, couple to the low-energy fermions that comprise the flat bands. These fluctuations are peaked at zero wave-vector, reflecting the "ferromagnetic-like" SU(4) ground state recently found in strong-coupling solutions of microscopic models for TBG. Focusing on electronic states related to symmetry-imposed points of the Fermi surface, dubbed here "valley hot-spots" and "van Hove hot-spots", we find that the coupling to the itinerant electrons partially lifts the huge degeneracy of the ferro-SU(4) ground state manifold, favoring inter-valley order, spin-valley coupled order, ferromagnetic order, spin-current order, and valley-polarized order, depending on details of the band structure. These fluctuations, in turn, promote attractive pairing interactions in a variety of closely competing channels, including a nodeless $f$-wave state, a nodal $i$-wave state, and topological $d+id$ and $p+ip$ states with unusual Chern numbers $2$ and $4$, respectively. Nematic superconductivity, although not realized as a primary instability of the system, still appears as a consequence of the near-degeneracy of superconducting order parameters that transform as one-dimensional and two-dimensional irreducible representations of the point group $D_{6}$.

cond-mat.supr-con↗

arXiv2020

Review of Text Style Transfer Based on Deep Learning

Text style transfer is a hot issue in recent natural language processing,which mainly studies the text to adapt to different specific situations, audiences and purposes by making some changes. The style of the text usually includes many aspects such as morphology, grammar, emotion, complexity, fluency, tense, tone and so on. In the traditional text style transfer model, the text style is generally relied on by experts knowledge and hand-designed rules, but with the application of deep learning in the field of natural language processing, the text style transfer method based on deep learning Started to be heavily researched. In recent years, text style transfer is becoming a hot issue in natural language processing research. This article summarizes the research on the text style transfer model based on deep learning in recent years, and summarizes, analyzes and compares the main research directions and progress. In addition, the article also introduces public data sets and evaluation indicators commonly used for text style transfer. Finally, the existing characteristics of the text style transfer model are summarized, and the future development trend of the text style transfer model based on deep learning is analyzed and forecasted.

cs.CL↗

arXiv2020

A topological Josephson junction platform for creating, manipulating, and braiding Majorana bound states

As part of the intense effort towards identifying platforms in which Majorana bound states can be realized and manipulated to perform qubit operations, we propose a topological Josephson junction architecture that achieves these capabilities and which can be experimentally implemented. The platform uses conventional superconducting electrodes deposited on a topological insulator film to form networks of proximity-coupled lateral Josephson junctions. Magnetic fields threading the network of junction barriers create Josephson vortices that host Majorana bound states localized in the junction where the local phase difference is an odd multiple of $π$, i.e. attached to the cores of the Josephson vortices. This enables us to manipulate the Majorana states by moving the Josephson vortices, achieving functionality exclusive to these systems in contrast to others, such as those composed of topological superconductor nanowires. We describe protocols for: 1) braiding localized Majorana states by exchange, 2) controlling the separation and hence the coupling of adjacent localized Majorana states to effect non-Abelian rotations via hybridization of the Majorana modes, and 3) reading out changes in the non-local parity correlations induced by such operations. These schemes make use of the application of current pulses and local magnetic field pulses to control the location of vortices, and measurements of the Josephson current-phase relation to reveal the presence of the Majorana bound states. We describe the architecture and schemes in the context of experiments currently underway.

cond-mat.mes-hall↗

arXiv2020

Efficient computation of two-loop amplitudes for Higgs boson pair production

We present a precise and efficient computation of the two-loop amplitudes entering the Higgs boson pair production process via gluon fusion. Our approach is based on the small-Higgs-mass expansion while keeping the full dependence on the top quark mass and other kinematic invariants. We compare our results to the up-to-date predictions based on a combination of sector decomposition and high-energy expansion. We find that our method provides precision numeric predictions in the entire phase space, while at the same time is highly efficient as the computation can be easily performed on a normal desktop or laptop computer. Our method is valuable for practical phenomenological studies of the Higgs boson pair production process, and can also be applied to other similar processes.

hep-ph↗

arXiv2020

Large-Scale MIDI-based Composer Classification

Music classification is a task to classify a music piece into labels such as genres or composers. We propose large-scale MIDI based composer classification systems using GiantMIDI-Piano, a transcription-based dataset. We propose to use piano rolls, onset rolls, and velocity rolls as input representations and use deep neural networks as classifiers. To our knowledge, we are the first to investigate the composer classification problem with up to 100 composers. By using convolutional recurrent neural networks as models, our MIDI based composer classification system achieves a 10-composer and a 100-composer classification accuracies of 0.648 and 0.385 (evaluated on 30-second clips) and 0.739 and 0.489 (evaluated on music pieces), respectively. Our MIDI based composer system outperforms several audio-based baseline classification systems, indicating the effectiveness of using compact MIDI representations for composer classification.

cs.SD↗

arXiv2020

Noise Robust TTS for Low Resource Speakers using Pre-trained Model and Speech Enhancement

With the popularity of deep neural network, speech synthesis task has achieved significant improvements based on the end-to-end encoder-decoder framework in the recent days. More and more applications relying on speech synthesis technology have been widely used in our daily life. Robust speech synthesis model depends on high quality and customized data which needs lots of collecting efforts. It is worth investigating how to take advantage of low-quality and low resource voice data which can be easily obtained from the Internet for usage of synthesizing personalized voice. In this paper, the proposed end-to-end speech synthesis model uses both speaker embedding and noise representation as conditional inputs to model speaker and noise information respectively. Firstly, the speech synthesis model is pre-trained with both multi-speaker clean data and noisy augmented data; then the pre-trained model is adapted on noisy low-resource new speaker data; finally, by setting the clean speech condition, the model can synthesize the new speaker's clean voice. Experimental results show that the speech generated by the proposed approach has better subjective evaluation results than the method directly fine-tuning pre-trained multi-speaker speech synthesis model with denoised new speaker data.

eess.AS↗

arXiv2020

Yukawa-SYK model and Self-tuned Quantum Criticality

Non-Fermi liquids (NFL) are a class of strongly interacting gapless fermionic systems without long-lived quasiparticle excitations. An important group of NFL model features itinerant fermions coupled to soft bosonic fluctuations near a quantum-critical point (QCP), and are widely believed to capture the essential physics of many unconventional superconductors. However numerically the direct observation of a canonical NFL behavior in such systems, characterized by a power-law form in the Green's function, has been elusive. Here we consider a Sachdev-Ye-Kitaev (SYK)-like model with random Yukawa interaction between critical bosons and fermions (dubbed Yukawa-SYK model). We show it is immune from minus-sign problem and hence can be solved exactly via large-scale quantum Monte Carlo simulation beyond the large-$N$ limit accessible to analytical approaches. Our simulation demonstrates the Yukawa-SYK model features "self-tuned quantum criticality", namely the system is critical independent of the bosonic bare mass. We put these results to test at finite $N$, and our unbiased numerics reveal clear evidence of these exotic quantum-critical NFL properties -- the power-law behavior in Green's function of fermions and bosons -- which propels the theoretical understanding of critical Planckian metals and unconventional superconductors.

cond-mat.str-el↗

arXiv2020

PANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognition

Audio pattern recognition is an important research topic in the machine learning area, and includes several tasks such as audio tagging, acoustic scene classification, music classification, speech emotion classification and sound event detection. Recently, neural networks have been applied to tackle audio pattern recognition problems. However, previous systems are built on specific datasets with limited durations. Recently, in computer vision and natural language processing, systems pretrained on large-scale datasets have generalized well to several tasks. However, there is limited research on pretraining systems on large-scale datasets for audio pattern recognition. In this paper, we propose pretrained audio neural networks (PANNs) trained on the large-scale AudioSet dataset. These PANNs are transferred to other audio related tasks. We investigate the performance and computational complexity of PANNs modeled by a variety of convolutional neural networks. We propose an architecture called Wavegram-Logmel-CNN using both log-mel spectrogram and waveform as input feature. Our best PANN system achieves a state-of-the-art mean average precision (mAP) of 0.439 on AudioSet tagging, outperforming the best previous system of 0.392. We transfer PANNs to six audio pattern recognition tasks, and demonstrate state-of-the-art performance in several of those tasks. We have released the source code and pretrained models of PANNs: https://github.com/qiuqiangkong/audioset_tagging_cnn.

cs.SD↗