arXiv subjects

Yi Zhou

Publications and source records attributed to Yi Zhou.

At least 361 records · Page 20Linked to original sources

arXiv2022

TTS-Guided Training for Accent Conversion Without Parallel Data

Accent Conversion (AC) seeks to change the accent of speech from one (source) to another (target) while preserving the speech content and speaker identity. However, many AC approaches rely on source-target parallel speech data. We propose a novel accent conversion framework without the need of parallel data. Specifically, a text-to-speech (TTS) system is first pretrained with target-accented speech data. This TTS model and its hidden representations are expected to be associated only with the target accent. Then, a speech encoder is trained to convert the accent of the speech under the supervision of the pretrained TTS model. In doing so, the source-accented speech and its corresponding transcription are forwarded to the speech encoder and the pretrained TTS, respectively. The output of the speech encoder is optimized to be the same as the text embedding in the TTS system. At run-time, the speech encoder is combined with the pretrained TTS decoder to convert the source-accented speech toward the target. In the experiments, we converted English with two source accents (Chinese and Indian) to the target accent (American/British/Canadian). Both objective metrics and subjective listening tests successfully validate that, without any parallel data, the proposed approach generates speech samples that are close to the target accent with high speech quality.

eess.AS↗

arXiv2022

Blow-up and lifespan estimate to a nonlinear wave equation in Schwarzschild spacetime

We study the semilinear wave equation with power type nonlinearity and small initial data in Schwarzschild spacetime. If the nonlinear exponent $p$ satisfies $2\le p<1+\sqrt 2$, we establish the sharp upper bound of lifespan estimate, while for the most delicate critical power $p=1+\sqrt2$, we show that the lifespan satisfies \[ T(\e)\le \exp\left(C\e^{-(2+\sqrt 2)}\right), \] the optimality of which remains to be proved. The key novelty is that the compact support of the initial data can be close to the event horizon. By combining the global existence result for $p>1+\sqrt 2$ obtained by Lindblad et al.(Math. Ann. 2014), we then give a positive answer to the interesting question posed by Dafermos and Rodnianski(J. Math. Pures Appl. 2005, the end of the first paragraph in page $1151$): $p=1+\sqrt 2$ is exactly the critical power of $p$ separating stability and blow-up.

math.AP↗

arXiv2022

Feature Aggregation and Propagation Network for Camouflaged Object Detection

Camouflaged object detection (COD) aims to detect/segment camouflaged objects embedded in the environment, which has attracted increasing attention over the past decades. Although several COD methods have been developed, they still suffer from unsatisfactory performance due to the intrinsic similarities between the foreground objects and background surroundings. In this paper, we propose a novel Feature Aggregation and Propagation Network (FAP-Net) for camouflaged object detection. Specifically, we propose a Boundary Guidance Module (BGM) to explicitly model the boundary characteristic, which can provide boundary-enhanced features to boost the COD performance. To capture the scale variations of the camouflaged objects, we propose a Multi-scale Feature Aggregation Module (MFAM) to characterize the multi-scale information from each layer and obtain the aggregated feature representations. Furthermore, we propose a Cross-level Fusion and Propagation Module (CFPM). In the CFPM, the feature fusion part can effectively integrate the features from adjacent layers to exploit the cross-level correlations, and the feature propagation part can transmit valuable context information from the encoder to the decoder network via a gate unit. Finally, we formulate a unified and end-to-end trainable framework where cross-level features can be effectively fused and propagated for capturing rich context information. Extensive experiments on three benchmark camouflaged datasets demonstrate that our FAP-Net outperforms other state-of-the-art COD models. Moreover, our model can be extended to the polyp segmentation task, and the comparison results further validate the effectiveness of the proposed model in segmenting polyps. The source code and results will be released at https://github.com/taozh2017/FAPNet.

cs.CV↗

arXiv2022

A Repulsive Force Unit for Garment Collision Handling in Neural Networks

Despite recent success, deep learning-based methods for predicting 3D garment deformation under body motion suffer from interpenetration problems between the garment and the body. To address this problem, we propose a novel collision handling neural network layer called Repulsive Force Unit (ReFU). Based on the signed distance function (SDF) of the underlying body and the current garment vertex positions, ReFU predicts the per-vertex offsets that push any interpenetrating vertex to a collision-free configuration while preserving the fine geometric details. We show that ReFU is differentiable with trainable parameters and can be integrated into different network backbones that predict 3D garment deformations. Our experiments show that ReFU significantly reduces the number of collisions between the body and the garment and better preserves geometric details compared to prior methods based on collision loss or post-processing optimization.

cs.GR↗

arXiv2022

On the Curious Case of $\ell_2$ norm of Sense Embeddings

We show that the $\ell_2$ norm of a static sense embedding encodes information related to the frequency of that sense in the training corpus used to learn the sense embeddings. This finding can be seen as an extension of a previously known relationship for word embeddings to sense embeddings. Our experimental results show that, in spite of its simplicity, the $\ell_2$ norm of sense embeddings is a surprisingly effective feature for several word sense related tasks such as (a) most frequent sense prediction, (b) Word-in-Context (WiC), and (c) Word Sense Disambiguation (WSD). In particular, by simply including the $\ell_2$ norm of a sense embedding as a feature in a classifier, we show that we can improve WiC and WSD methods that use static sense embeddings.

cs.CL↗

arXiv2022

Two-stage superconductivity in the Hatsugai-Kohmoto-BCS model

Superconductivity in strongly correlated electrons can emerge out from a normal state that is beyond the Landau's Fermi liquid paradigm, often dubbed as "non-Fermi liquid". While the theory for non-Fermi liquid is still not yet conclusive, a recent study on the exactly-solvable Hatsugai-Kohmoto (HK) model has suggested a non-Fermi liquid ground state whose Green's function resembles the Yang-Rice-Zhang ansatz for cuprates [P. W. Phillips, L. Yeo and E. W. Huang, Nat. Phys. $\bf{16}$, 1175 (2020)]. Similar to the effect of on-site Coulomb repulsion in the Hubbard model, the repulsive interaction in the HK model divides the momentum space into three parts: empty, single-occupied and double-occupied regions, that are separated from each other by two distinct Fermi surfaces. In the presence of an additional Bardeen-Cooper-Schrieffer (BCS)-type pairing interaction of a moderate strength, we show that the system exhibits a "two-stage superconductivity" feature as temperature decreases: a first-order superconducting transition occurs at a temperature $T_{\rm c}$ that is followed by a sudden increase of the superconducting order parameter at a lower temperature $T_{\rm c}^{\prime}<T_{\rm c}$. At the first stage, $T_{\rm c}^{\prime}<T<T_{\rm c}$, the pairing function arises and the entropy is released only in the vicinity of the two Fermi surfaces; while at the second stage, $T<T_{\rm c}^{\prime}$, the pairing function becomes significant and the entropy is further released in deep (single-occupied) region in the Fermi sea. The phase transitions are analyzed within the Ginzburg-Landau theory. Our work sheds new light on unconventional superconductivity in strongly correlated electrons.

cond-mat.supr-con↗

arXiv2022

Interplay between Pair Density Wave and a Nested Fermi Surface

We show that spontaneous time-reversal-symmetry (TRS) breaking can naturally arise from the interplay between pair density wave (PDW) ordering at multiple momenta and nesting of Fermi surfaces (FS). Concretely, we consider pair-density-wave superconductivity on a hexagonal lattice with nested FS at $3/4$ electron filling, which is related to a recently discovered superconductor CsV$_3$Sb$_5$. Because of nesting of the FS, each momentum $\bk$ on the FS has at least two counterparts $-\bk\pm\mathbf{Q}_α$ ($α=1,2,3$) on the FS to form finite momentum ($\pm\mathbf{Q}_α$) Cooper pairs, resulting in a TRS and inversion broken PDW state with stable Bogoliubov Fermi pockets. Various spectra, including (local) density of states, electron spectral function and the effect of quasi-particle interference, have been investigated. The partial melting of the PDW will give rise to $4\times{}4$ and $\frac{4}{\sqrt{3}}\times\frac{4}{\sqrt{3}}$ CDW orders, in addition to the $2\times2$ CDW. Possible implications to real materials such as CsV$_3$Sb$_5$ and future experiments have been further discussed.

cond-mat.supr-con↗

arXiv2022

NeMF: Neural Motion Fields for Kinematic Animation

We present an implicit neural representation to learn the spatio-temporal space of kinematic motions. Unlike previous work that represents motion as discrete sequential samples, we propose to express the vast motion space as a continuous function over time, hence the name Neural Motion Fields (NeMF). Specifically, we use a neural network to learn this function for miscellaneous sets of motions, which is designed to be a generative model conditioned on a temporal coordinate $t$ and a random vector $z$ for controlling the style. The model is then trained as a Variational Autoencoder (VAE) with motion encoders to sample the latent space. We train our model with a diverse human motion dataset and quadruped dataset to prove its versatility, and finally deploy it as a generic motion prior to solve task-agnostic problems and show its superiority in different motion generation and editing applications, such as motion interpolation, in-betweening, and re-navigating. More details can be found on our project page: https://cs.yale.edu/homes/che/projects/nemf/.

cs.CV↗

arXiv2022

PARAGEN : A Parallel Generation Toolkit

PARAGEN is a PyTorch-based NLP toolkit for further development on parallel generation. PARAGEN provides thirteen types of customizable plugins, helping users to experiment quickly with novel ideas across model architectures, optimization, and learning strategies. We implement various features, such as unlimited data loading and automatic model selection, to enhance its industrial usage. ParaGen is now deployed to support various research and industry applications at ByteDance. PARAGEN is available at https://github.com/bytedance/ParaGen.

cs.CL↗

arXiv2022

Zero-Shot 3D Drug Design by Sketching and Generating

Drug design is a crucial step in the drug discovery cycle. Recently, various deep learning-based methods design drugs by generating novel molecules from scratch, avoiding traversing large-scale drug libraries. However, they depend on scarce experimental data or time-consuming docking simulation, leading to overfitting issues with limited training data and slow generation speed. In this study, we propose the zero-shot drug design method DESERT (Drug dEsign by SkEtching and geneRaTing). Specifically, DESERT splits the design process into two stages: sketching and generating, and bridges them with the molecular shape. The two-stage fashion enables our method to utilize the large-scale molecular database to reduce the need for experimental data and docking simulation. Experiments show that DESERT achieves a new state-of-the-art at a fast speed.

cs.CE↗

arXiv2022

Federated XGBoost on Sample-Wise Non-IID Data

Federated Learning (FL) is a paradigm for jointly training machine learning algorithms in a decentralized manner which allows for parties to communicate with an aggregator to create and train a model, without exposing the underlying raw data distribution of the local parties involved in the training process. Most research in FL has been focused on Neural Network-based approaches, however Tree-Based methods, such as XGBoost, have been underexplored in Federated Learning due to the challenges in overcoming the iterative and additive characteristics of the algorithm. Decision tree-based models, in particular XGBoost, can handle non-IID data, which is significant for algorithms used in Federated Learning frameworks since the underlying characteristics of the data are decentralized and have risks of being non-IID by nature. In this paper, we focus on investigating the effects of how Federated XGBoost is impacted by non-IID distributions by performing experiments on various sample size-based data skew scenarios and how these models perform under various non-IID scenarios. We conduct a set of extensive experiments across multiple different datasets and different data skew partitions. Our experimental results demonstrate that despite the various partition ratios, the performance of the models stayed consistent and performed close to or equally well against models that were trained in a centralized manner.

cs.LG↗

arXiv2022

Generalized persistence of entropy weak solutions for system of hyperbolic conservation laws

Let $u(t,x)$ be the solution to the Cauchy problem of a scalar conservation law in one space dimension. It is well known that even for smooth initial data the solution can become discontinuous in finite time and global entropy weak solution can best lie in the space of bounded total variations. It is impossible that the solutions belong to ,for example ,$H^1$ because by Sobolev embedding theorem $H^1$ functions are H$\mathrm{\ddot{o}}$lder continuous. However, we note that from any point $(t,x)$ we can draw a generalized characteristic downward which meets the initial axis at $y=α(t,x)$. if we regard $u$ as a function of $(t,y)$, it indeed belongs to $H^1$ as a function of $y$ if the initial data belongs to $H^1$. We may call this generalized persistence (of high regularity) of the entropy weak solutions. The main purpose of this paper is to prove some kinds of generalized persistence (of high regularity) for the scalar and $2\times 2$ Temple system of hyperbolic conservation laws in one space dimension .

math.AP↗

arXiv2022

Learning Visibility for Robust Dense Human Body Estimation

Estimating 3D human pose and shape from 2D images is a crucial yet challenging task. While prior methods with model-based representations can perform reasonably well on whole-body images, they often fail when parts of the body are occluded or outside the frame. Moreover, these results usually do not faithfully capture the human silhouettes due to their limited representation power of deformable models (e.g., representing only the naked body). An alternative approach is to estimate dense vertices of a predefined template body in the image space. Such representations are effective in localizing vertices within an image but cannot handle out-of-frame body parts. In this work, we learn dense human body estimation that is robust to partial observations. We explicitly model the visibility of human joints and vertices in the x, y, and z axes separately. The visibility in x and y axes help distinguishing out-of-frame cases, and the visibility in depth axis corresponds to occlusions (either self-occlusions or occlusions by other objects). We obtain pseudo ground-truths of visibility labels from dense UV correspondences and train a neural network to predict visibility along with 3D coordinates. We show that visibility can serve as 1) an additional signal to resolve depth ordering ambiguities of self-occluded vertices and 2) a regularization term when fitting a human body model to the predictions. Extensive experiments on multiple 3D human datasets demonstrate that visibility modeling significantly improves the accuracy of human body estimation, especially for partial-body cases. Our project page with code is at: https://github.com/chhankyao/visdb.

cs.CV↗

arXiv2022

Extracting Densest Sub-hypergraph with Convex Edge-weight Functions

The densest subgraph problem (DSG) aiming at finding an induced subgraph such that the average edge-weights of the subgraph is maximized, is a well-studied problem. However, when the input graph is a hypergraph, the existing notion of DSG fails to capture the fact that a hyperedge partially belonging to an induced sub-hypergraph is also a part of the sub-hypergraph. To resolve the issue, we suggest a function $f_e:\mathbb{Z}_{\ge0}\rightarrow \mathbb{R}_{\ge 0}$ to represent the partial edge-weight of a hyperedge $e$ in the input hypergraph $\mathcal{H}=(V,\mathcal{E},f)$ and formulate a generalized densest sub-hypergraph problem (GDSH) as $\max_{S\subseteq V}\frac{\sum_{e\in \mathcal{E}}{f_e(|e\cap S|)}}{|S|}$. We demonstrate that, when all the edge-weight functions are non-decreasing convex, GDSH can be solved in polynomial-time by the linear program-based algorithm, the network flow-based algorithm and the greedy $\frac{1}{r}$-approximation algorithm where $r$ is the rank of the input hypergraph. Finally, we investigate the computational tractability of GDSH where some edge-weight functions are non-convex.

cs.DS↗

arXiv2022

DeTrust-FL: Privacy-Preserving Federated Learning in Decentralized Trust Setting

Federated learning has emerged as a privacy-preserving machine learning approach where multiple parties can train a single model without sharing their raw training data. Federated learning typically requires the utilization of multi-party computation techniques to provide strong privacy guarantees by ensuring that an untrusted or curious aggregator cannot obtain isolated replies from parties involved in the training process, thereby preventing potential inference attacks. Until recently, it was thought that some of these secure aggregation techniques were sufficient to fully protect against inference attacks coming from a curious aggregator. However, recent research has demonstrated that a curious aggregator can successfully launch a disaggregation attack to learn information about model updates of a target party. This paper presents DeTrust-FL, an efficient privacy-preserving federated learning framework for addressing the lack of transparency that enables isolation attacks, such as disaggregation attacks, during secure aggregation by assuring that parties' model updates are included in the aggregated model in a private and secure manner. DeTrust-FL proposes a decentralized trust consensus mechanism and incorporates a recently proposed decentralized functional encryption (FE) scheme in which all parties agree on a participation matrix before collaboratively generating decryption key fragments, thereby gaining control and trust over the secure aggregation process in a decentralized setting. Our experimental evaluation demonstrates that DeTrust-FL outperforms state-of-the-art FE-based secure multi-party aggregation solutions in terms of training time and reduces the volume of data transferred. In contrast to existing approaches, this is achieved without creating any trust dependency on external trusted entities.

cs.CR↗

arXiv2022

A Fast and Convergent Proximal Algorithm for Regularized Nonconvex and Nonsmooth Bi-level Optimization

Many important machine learning applications involve regularized nonconvex bi-level optimization. However, the existing gradient-based bi-level optimization algorithms cannot handle nonconvex or nonsmooth regularizers, and they suffer from a high computation complexity in nonconvex bi-level optimization. In this work, we study a proximal gradient-type algorithm that adopts the approximate implicit differentiation (AID) scheme for nonconvex bi-level optimization with possibly nonconvex and nonsmooth regularizers. In particular, the algorithm applies the Nesterov's momentum to accelerate the computation of the implicit gradient involved in AID. We provide a comprehensive analysis of the global convergence properties of this algorithm through identifying its intrinsic potential function. In particular, we formally establish the convergence of the model parameters to a critical point of the bi-level problem, and obtain an improved computation complexity $\mathcal{O}(κ^{3.5}ε^{-2})$ over the state-of-the-art result. Moreover, we analyze the asymptotic convergence rates of this algorithm under a class of local nonconvex geometries characterized by a Łojasiewicz-type gradient inequality. Experiment on hyper-parameter optimization demonstrates the effectiveness of our algorithm.

cs.LG↗

arXiv2022

Collective effects in an incompressible electronic liquid

Starting from the Landau's kinetic equation, we show that an electronic liquid in $d=2,3$ dimensions depicted by a Landau type effective theory will become incompressible on condition that the Landau parameters satisfy either (i) $1+F_{1}^{s}/d=0$ or (ii) $F_{0}^{s}\to{}+\infty$. The condition (i) is the Pomeranchuk instability in the current channel and suggests a quantum spin liquid (QSL) state with a spinon Fermi surface; while the condition (ii) means that the strong repulsion in the charge channel and leads to a conventional charge and thermal insulator. In the collisionless regime ($ωτ\gg{}1$) and the hydrodynamic regime ($ωτ\ll{}1$), the zero and first sound modes have been studied and classified by symmetries, including the longitudinal and transverse modes in $d=2,3$ and the higher angular momentum modes in $d=3$. The sufficient (and/or necessary) conditions of these collective modes have been revealed. It has been demonstrated that some of these collective modes will behave in quite different manners under two incompressibility conditions (i) or (ii). Possible nematic QSL states and a hierarchy structure for gapless QSL states have been proposed in $d=3$.

cond-mat.str-el↗

arXiv2022

Accelerated Proximal Alternating Gradient-Descent-Ascent for Nonconvex Minimax Machine Learning

Alternating gradient-descent-ascent (AltGDA) is an optimization algorithm that has been widely used for model training in various machine learning applications, which aims to solve a nonconvex minimax optimization problem. However, the existing studies show that it suffers from a high computation complexity in nonconvex minimax optimization. In this paper, we develop a single-loop and fast AltGDA-type algorithm that leverages proximal gradient updates and momentum acceleration to solve regularized nonconvex minimax optimization problems. By leveraging the momentum acceleration technique, we prove that the algorithm converges to a critical point in nonconvex minimax optimization and achieves a computation complexity in the order of $\mathcal{O}(κ^{\frac{11}{6}}ε^{-2})$, where $ε$ is the desired level of accuracy and $κ$ is the problem's condition number. {Such a computation complexity improves the state-of-the-art complexities of single-loop GDA and AltGDA algorithms (see the summary of comparison in \Cref{table1})}. We demonstrate the effectiveness of our algorithm via an experiment on adversarial deep learning.

cs.LG↗