Searcharxiv⌕ Search

arXiv subjects

Chen Zhang

Publications and source records attributed to Chen Zhang.

At least 487 records · Page 27Linked to original sources
arXiv2022

Structural Bias for Aspect Sentiment Triplet Extraction

Structural bias has recently been exploited for aspect sentiment triplet extraction (ASTE) and led to improved performance. On the other hand, it is recognized that explicitly incorporating structural bias would have a negative impact on efficiency, whereas pretrained language models (PLMs) can already capture implicit structures. Thus, a natural question arises: Is structural bias still a necessity in the context of PLMs? To answer the question, we propose to address the efficiency issues by using an adapter to integrate structural bias in the PLM and using a cheap-to-compute relative position structure in place of the syntactic dependency structure. Benchmarking evaluation is conducted on the SemEval datasets. The results show that our proposed structural adapter is beneficial to PLMs and achieves state-of-the-art performance over a range of strong baselines, yet with a light parameter demand and low latency. Meanwhile, we give rise to the concern that the current evaluation default with data of small scale is under-confident. Consequently, we release a large-scale dataset for ASTE. The results on the new dataset hint that the structural adapter is confidently effective and efficient to a large scale. Overall, we draw the conclusion that structural bias shall still be a necessity even with PLMs.

cs.CL↗
arXiv2022

ANT: Exploiting Adaptive Numerical Data Type for Low-bit Deep Neural Network Quantization

Quantization is a technique to reduce the computation and memory cost of DNN models, which are getting increasingly large. Existing quantization solutions use fixed-point integer or floating-point types, which have limited benefits, as both require more bits to maintain the accuracy of original models. On the other hand, variable-length quantization uses low-bit quantization for normal values and high-precision for a fraction of outlier values. Even though this line of work brings algorithmic benefits, it also introduces significant hardware overheads due to variable-length encoding and decoding. In this work, we propose a fixed-length adaptive numerical data type called ANT to achieve low-bit quantization with tiny hardware overheads. Our data type ANT leverages two key innovations to exploit the intra-tensor and inter-tensor adaptive opportunities in DNN models. First, we propose a particular data type, flint, that combines the advantages of float and int for adapting to the importance of different values within a tensor. Second, we propose an adaptive framework that selects the best type for each tensor according to its distribution characteristics. We design a unified processing element architecture for ANT and show its ease of integration with existing DNN accelerators. Our design results in 2.8$\times$ speedup and 2.5$\times$ energy efficiency improvement over the state-of-the-art quantization accelerators.

cs.LG↗
arXiv2022

Connection-oriented and Connectionless Quantum Internet Considering Quantum Repeaters

With the rapid development of quantum information and technology in recent years, the construction of quantum internet for interconnecting all kinds of quantum devices, such as quantum processors and sensors, will be the next trend for practical quantum applications. In this paper, we propose the protocols for construction of connection-oriented and connectionless quantum networks by considering the concrete quantum repeater (QR) nodes. Four classes of QRs networks are considered first and designed with two types of protocols in link layer, i.e. simultaneous and one-by-one link. Based on those two link models, the connection-oriented protocol is presented for all classes of QRs networks and the connectionless protocol is proposed for the first, second and third classes QRs networks by only one-by-one link. Furthermore, we introduce a new hybrid connection model of quantum networks combined with connection-oriented and connectionless for practical uses. Our work is a new attempt to study the model of the network layer for different kinds of QR networks and paves the way for developing the protocol stack of universal large-scale quantum internet.

quant-ph↗
arXiv2022

Aspect-specific Context Modeling for Aspect-based Sentiment Analysis

Aspect-based sentiment analysis (ABSA) aims at predicting sentiment polarity (SC) or extracting opinion span (OE) expressed towards a given aspect. Previous work in ABSA mostly relies on rather complicated aspect-specific feature induction. Recently, pretrained language models (PLMs), e.g., BERT, have been used as context modeling layers to simplify the feature induction structures and achieve state-of-the-art performance. However, such PLM-based context modeling can be not that aspect-specific. Therefore, a key question is left under-explored: how the aspect-specific context can be better modeled through PLMs? To answer the question, we attempt to enhance aspect-specific context modeling with PLM in a non-intrusive manner. We propose three aspect-specific input transformations, namely aspect companion, aspect prompt, and aspect marker. Informed by these transformations, non-intrusive aspect-specific PLMs can be achieved to promote the PLM to pay more attention to the aspect-specific context in a sentence. Additionally, we craft an adversarial benchmark for ABSA (advABSA) to see how aspect-specific modeling can impact model robustness. Extensive experimental results on standard and adversarial benchmarks for SC and OE demonstrate the effectiveness and robustness of the proposed method, yielding new state-of-the-art performance on OE and competitive performance on SC.

cs.CL↗
arXiv2022

ReLyMe: Improving Lyric-to-Melody Generation by Incorporating Lyric-Melody Relationships

Lyric-to-melody generation, which generates melody according to given lyrics, is one of the most important automatic music composition tasks. With the rapid development of deep learning, previous works address this task with end-to-end neural network models. However, deep learning models cannot well capture the strict but subtle relationships between lyrics and melodies, which compromises the harmony between lyrics and generated melodies. In this paper, we propose ReLyMe, a method that incorporates Relationships between Lyrics and Melodies from music theory to ensure the harmony between lyrics and melodies. Specifically, we first introduce several principles that lyrics and melodies should follow in terms of tone, rhythm, and structure relationships. These principles are then integrated into neural network lyric-to-melody models by adding corresponding constraints during the decoding process to improve the harmony between lyrics and melodies. We use a series of objective and subjective metrics to evaluate the generated melodies. Experiments on both English and Chinese song datasets show the effectiveness of ReLyMe, demonstrating the superiority of incorporating lyric-melody relationships from the music domain into neural lyric-to-melody generation.

cs.SD↗
arXiv2022

SummScore: A Comprehensive Evaluation Metric for Summary Quality Based on Cross-Encoder

Text summarization models are often trained to produce summaries that meet human quality requirements. However, the existing evaluation metrics for summary text are only rough proxies for summary quality, suffering from low correlation with human scoring and inhibition of summary diversity. To solve these problems, we propose SummScore, a comprehensive metric for summary quality evaluation based on CrossEncoder. Firstly, by adopting the original-summary measurement mode and comparing the semantics of the original text, SummScore gets rid of the inhibition of summary diversity. With the help of the text-matching pre-training Cross-Encoder, SummScore can effectively capture the subtle differences between the semantics of summaries. Secondly, to improve the comprehensiveness and interpretability, SummScore consists of four fine-grained submodels, which measure Coherence, Consistency, Fluency, and Relevance separately. We use semi-supervised multi-rounds of training to improve the performance of our model on extremely limited annotated data. Extensive experiments show that SummScore significantly outperforms existing evaluation metrics in the above four dimensions in correlation with human scoring. We also provide the quality evaluation results of SummScore on 16 mainstream summarization models for later research.

cs.CL↗
arXiv2022

BANet: Motion Forecasting with Boundary Aware Network

We propose a motion forecasting model called BANet, which means Boundary-Aware Network, and it is a variant of LaneGCN. We believe that it is not enough to use only the lane centerline as input to obtain the embedding features of the vector map nodes. The lane centerline can only provide the topology of the lanes, and other elements of the vector map also contain rich information. For example, the lane boundary can provide traffic rule constraint information such as whether it is possible to change lanes which is very important. Therefore, we achieved better performance by encoding more vector map elements in the motion forecasting model.We report our results on the 2022 Argoverse2 Motion Forecasting challenge and rank 1st on the test leaderboard.

cs.CV↗
arXiv2022

Adaptable Text Matching via Meta-Weight Regulator

Neural text matching models have been used in a range of applications such as question answering and natural language inference, and have yielded a good performance. However, these neural models are of a limited adaptability, resulting in a decline in performance when encountering test examples from a different dataset or even a different task. The adaptability is particularly important in the few-shot setting: in many cases, there is only a limited amount of labeled data available for a target dataset or task, while we may have access to a richly labeled source dataset or task. However, adapting a model trained on the abundant source data to a few-shot target dataset or task is challenging. To tackle this challenge, we propose a Meta-Weight Regulator (MWR), which is a meta-learning approach that learns to assign weights to the source examples based on their relevance to the target loss. Specifically, MWR first trains the model on the uniformly weighted source examples, and measures the efficacy of the model on the target examples via a loss function. By iteratively performing a (meta) gradient descent, high-order gradients are propagated to the source examples. These gradients are then used to update the weights of source examples, in a way that is relevant to the target performance. As MWR is model-agnostic, it can be applied to any backbone neural model. Extensive experiments are conducted with various backbone text matching models, on four widely used datasets and two tasks. The results demonstrate that our proposed approach significantly outperforms a number of existing adaptation methods and effectively improves the cross-dataset and cross-task adaptability of the neural text matching models in the few-shot setting.

cs.IR↗
arXiv2022

NaturalSpeech: End-to-End Text to Speech Synthesis with Human-Level Quality

Text to speech (TTS) has made rapid progress in both academia and industry in recent years. Some questions naturally arise that whether a TTS system can achieve human-level quality, how to define/judge that quality and how to achieve it. In this paper, we answer these questions by first defining the human-level quality based on the statistical significance of subjective measure and introducing appropriate guidelines to judge it, and then developing a TTS system called NaturalSpeech that achieves human-level quality on a benchmark dataset. Specifically, we leverage a variational autoencoder (VAE) for end-to-end text to waveform generation, with several key modules to enhance the capacity of the prior from text and reduce the complexity of the posterior from speech, including phoneme pre-training, differentiable duration modeling, bidirectional prior/posterior modeling, and a memory mechanism in VAE. Experiment evaluations on popular LJSpeech dataset show that our proposed NaturalSpeech achieves -0.01 CMOS (comparative mean opinion score) to human recordings at the sentence level, with Wilcoxon signed rank test at p-level p >> 0.05, which demonstrates no statistically significant difference from human recordings for the first time on this dataset.

eess.AS↗
arXiv2022

Onset of nonlinear electroosmotic flow under AC electric field

Nonlinearity of electroosmotic flows (EOFs) is ubiquitous and plays a crucial role in the mass and energy transfer in ion transport, specimen mixing, electrochemistry reaction, and electric energy storage and utilizing. When and how the transition from a linear regime to a nonlinear one is essential for understanding, prohibiting or utilizing nonlinear EOF. However, suffers the lacking of reliable experimental instruments with high spatial and temporal resolutions, the investigation of the onset of nonlinear EOF still stays in theory. Herein, we experimentally studied the velocity fluctuations of EOFs driven by AC electric field via ultra-sensitive fluorescent blinking tricks. The linear and nonlinear AC EOFs are successfully identified from both the time trace and energy spectra of velocity fluctuations. The critical electric field ($E_{A,C}$) separating the two statuses is determined and is discovered by defining a generalized scaling law with respect to the convection velocity ($U$) and AC frequency ($f_f$) as $E_{A,C}$~${f_f}^{0.48-0.027U}$. The universal control parameters are determined with surprising accuracy for governing the status of AC EOFs. We hope the current investigation could be essential in the development of both theory and applications of nonlinear EOF.

physics.flu-dyn↗
arXiv2022

Familiarity-based Collaborative Team Recognition in Academic Social Networks

Collaborative teamwork is key to major scientific discoveries. However, the prevalence of collaboration among researchers makes team recognition increasingly challenging. Previous studies have demonstrated that people are more likely to collaborate with individuals they are familiar with. In this work, we employ the definition of familiarity and then propose MOTO (faMiliarity-based cOllaborative Team recOgnition algorithm) to recognize collaborative teams. MOTO calculates the shortest distance matrix within the global collaboration network and the local density of each node. Central team members are initially recognized based on local density. Then MOTO recognizes the remaining team members by using the familiarity metric and shortest distance matrix. Extensive experiments have been conducted upon a large-scale data set. The experimental results show that compared with baseline methods, MOTO can recognize the largest number of teams. The teams recognized by MOTO possess more cohesive team structures and lower team communication costs compared with other methods. MOTO utilizes familiarity in team recognition to identify cohesive academic teams. The recognized teams are in line with real-world collaborative teamwork patterns. Based on team recognition using MOTO, the research team structure and performance are further analyzed for given time periods. The number of teams that consist of members from different institutions increases gradually. Such teams are found to perform better in comparison with those whose members are from the same institution.

cs.SI↗
arXiv2022

Split Hierarchical Variational Compression

Variational autoencoders (VAEs) have witnessed great success in performing the compression of image datasets. This success, made possible by the bits-back coding framework, has produced competitive compression performance across many benchmarks. However, despite this, VAE architectures are currently limited by a combination of coding practicalities and compression ratios. That is, not only do state-of-the-art methods, such as normalizing flows, often demonstrate out-performance, but the initial bits required in coding makes single and parallel image compression challenging. To remedy this, we introduce Split Hierarchical Variational Compression (SHVC). SHVC introduces two novelties. Firstly, we propose an efficient autoregressive prior, the autoregressive sub-pixel convolution, that allows a generalisation between per-pixel autoregressions and fully factorised probability models. Secondly, we define our coding framework, the autoregressive initial bits, that flexibly supports parallel coding and avoids -- for the first time -- many of the practicalities commonly associated with bits-back coding. In our experiments, we demonstrate SHVC is able to achieve state-of-the-art compression performance across full-resolution lossless image compression tasks, with up to 100x fewer model parameters than competing VAE approaches.

eess.IV↗
arXiv2022

Automatic Song Translation for Tonal Languages

This paper develops automatic song translation (AST) for tonal languages and addresses the unique challenge of aligning words' tones with melody of a song in addition to conveying the original meaning. We propose three criteria for effective AST -- preserving meaning, singability and intelligibility -- and design metrics for these criteria. We develop a new benchmark for English--Mandarin song translation and develop an unsupervised AST system, Guided AliGnment for Automatic Song Translation (GagaST), which combines pre-training with three decoding constraints. Both automatic and human evaluations show GagaST successfully balances semantics and singability.

cs.CL↗
arXiv2022

Angle-resolved photoemission spectroscopy study of the charge density wave order in layered semiconductor EuTe4

Layered tellurides have been extensively studied as a platform for investigating the Fermi surface (FS) nesting-driven charge density wave (CDW) states. EuTe4, one of quasi-two-dimensional (quasi-2D) binary rare-earth tetratellurides CDW compounds, with unconventional hysteretic transition, is currently receiving much attention. Here, the CDW modulation vector, momentum and temperature dependence of CDW gaps in EuTe4 are investigated using angle-resolved photoemission spectroscopy. Our results reveal that (i) a FS nesting vector q ~ 0.67 b* drives the formation of CDW state, (ii) a large anisotropic CDW gap is fully open in the whole FS, and maintains a considerable size even at 300 K, leading to appearance of semiconductor properties, (iii) an abnormal non-monotonic increase of CDW gap in magnitude as a function of temperature, (iv) an extra, larger gap opens at a higher binding energy due to the interaction between the different orbits of the main bands.

cond-mat.mtrl-sci↗
arXiv2022

Three-Port Impedance Model and Validation of VSCs for Stability Analysis

Modern power system is undergoing a paradigm shift from the synchronous generators-based system to the power electronics converters-dominated system. With the high penetration of converters, serious stability problems are provoked, especially the wideband oscillations. Various studies have been conducted in this respect, while most of them separate the ac-side stability with the dc-side stability. However, for the stability analysis of the hybrid AC/DC grid, it is necessary to consider the converter ac-side and dc-side, simultaneously. In this paper, the stability analysis of voltage source converters (VSCs) considering both ac and dc dynamics is carried out. At first, the three-port AC/DC admittance model of VSCs is established, and the corresponding measurement method from simulations is presented to validate its accuracy. Secondly, based on such three-port model, two stability analysis methods are presented: the one is based on the system open-loop model, where the stability can be judged via the Generalized Nyquist Criterion (GNC); the other one is based on the system closed-loop model, whose stability can be predicted through the pole-zero calculation. At last, a test AC/DC system is built in MATLAB/Simulink, by which the effectiveness of the three-port model-based stability analysis is validated.

physics.app-ph↗
arXiv2022

Data-Driven Decision Making in COVID-19 Response: A Survey

COVID-19 has spread all over the world, having an enormous effect on our daily life and work. In response to the epidemic, a lot of important decisions need to be taken to save communities and economies worldwide. Data clearly plays a vital role in effective decision making. Data-driven decision making uses data related evidence and insights to guide the decision making process and to verify the plan of action before it is committed. To better handle the epidemic, governments and policy making institutes have investigated abundant data originating from COVID-19. These data include those related to medicine, knowledge, media, etc. Based on these data, many prevention and control policies are made. In this survey paper, we summarize the progress of data-driven decision making in the response to COVID-19, including COVID-19 prevention and control, psychological counselling, financial aid, work resumption, and school re-opening. We also propose some current challenges and open issues in data-driven decision making, including data collection and quality, complex data analysis, and fairness in decision making. This survey paper sheds light on current policy making driven by data, which also provides a feasible direction for further scientific research.

cs.SI↗
arXiv2022

S3T: Self-Supervised Pre-training with Swin Transformer for Music Classification

In this paper, we propose S3T, a self-supervised pre-training method with Swin Transformer for music classification, aiming to learn meaningful music representations from massive easily accessible unlabeled music data. S3T introduces a momentum-based paradigm, MoCo, with Swin Transformer as its feature extractor to music time-frequency domain. For better music representations learning, S3T contributes a music data augmentation pipeline and two specially designed pre-processors. To our knowledge, S3T is the first method combining the Swin Transformer with a self-supervised learning method for music classification. We evaluate S3T on music genre classification and music tagging tasks with linear classifiers trained on learned representations. Experimental results show that S3T outperforms the previous self-supervised method (CLMR) by 12.5 percents top-1 accuracy and 4.8 percents PR-AUC on two tasks respectively, and also surpasses the task-specific state-of-the-art supervised methods. Besides, S3T shows advances in label efficiency using only 10% labeled data exceeding CLMR on both tasks with 100% labeled data.

eess.AS↗