arXiv subjects

Chao Zhang

Publications and source records attributed to Chao Zhang.

At least 775 records · Page 43Linked to original sources

arXiv2022

UnfoldML: Cost-Aware and Uncertainty-Based Dynamic 2D Prediction for Multi-Stage Classification

Machine Learning (ML) research has focused on maximizing the accuracy of predictive tasks. ML models, however, are increasingly more complex, resource intensive, and costlier to deploy in resource-constrained environments. These issues are exacerbated for prediction tasks with sequential classification on progressively transitioned stages with ''happens-before'' relation between them.We argue that it is possible to ''unfold'' a monolithic single multi-class classifier, typically trained for all stages using all data, into a series of single-stage classifiers. Each single-stage classifier can be cascaded gradually from cheaper to more expensive binary classifiers that are trained using only the necessary data modalities or features required for that stage. UnfoldML is a cost-aware and uncertainty-based dynamic 2D prediction pipeline for multi-stage classification that enables (1) navigation of the accuracy/cost tradeoff space, (2) reducing the spatio-temporal cost of inference by orders of magnitude, and (3) early prediction on proceeding stages. UnfoldML achieves orders of magnitude better cost in clinical settings, while detecting multi-stage disease development in real time. It achieves within 0.1% accuracy from the highest-performing multi-class baseline, while saving close to 20X on spatio-temporal cost of inference and earlier (3.5hrs) disease onset prediction. We also show that UnfoldML generalizes to image classification, where it can predict different level of labels (from coarse to fine) given different level of abstractions of a image, saving close to 5X cost with as little as 0.4% accuracy reduction.

cs.LG↗

arXiv2022

RoChBert: Towards Robust BERT Fine-tuning for Chinese

Despite of the superb performance on a wide range of tasks, pre-trained language models (e.g., BERT) have been proved vulnerable to adversarial texts. In this paper, we present RoChBERT, a framework to build more Robust BERT-based models by utilizing a more comprehensive adversarial graph to fuse Chinese phonetic and glyph features into pre-trained representations during fine-tuning. Inspired by curriculum learning, we further propose to augment the training dataset with adversarial texts in combination with intermediate samples. Extensive experiments demonstrate that RoChBERT outperforms previous methods in significant ways: (i) robust -- RoChBERT greatly improves the model robustness without sacrificing accuracy on benign texts. Specifically, the defense lowers the success rates of unlimited and limited attacks by 59.43% and 39.33% respectively, while remaining accuracy of 93.30%; (ii) flexible -- RoChBERT can easily extend to various language models to solve different downstream tasks with excellent performance; and (iii) efficient -- RoChBERT can be directly applied to the fine-tuning stage without pre-training language model from scratch, and the proposed data augmentation method is also low-cost.

cs.CL↗

arXiv2022

ReSel: N-ary Relation Extraction from Scientific Text and Tables by Learning to Retrieve and Select

We study the problem of extracting N-ary relation tuples from scientific articles. This task is challenging because the target knowledge tuples can reside in multiple parts and modalities of the document. Our proposed method ReSel decomposes this task into a two-stage procedure that first retrieves the most relevant paragraph/table and then selects the target entity from the retrieved component. For the high-level retrieval stage, ReSel designs a simple and effective feature set, which captures multi-level lexical and semantic similarities between the query and components. For the low-level selection stage, ReSel designs a cross-modal entity correlation graph along with a multi-view architecture, which models both semantic and document-structural relations between entities. Our experiments on three scientific information extraction datasets show that ReSel outperforms state-of-the-art baselines significantly.

cs.CL↗

arXiv2022

Pronunciation Generation for Foreign Language Words in Intra-Sentential Code-Switching Speech Recognition

Code-Switching refers to the phenomenon of switching languages within a sentence or discourse. However, limited code-switching , different language phoneme-sets and high rebuilding costs throw a challenge to make the specialized acoustic model for code-switching speech recognition. In this paper, we make use of limited code-switching data as driving materials and explore a shortcut to quickly develop intra-sentential code-switching recognition skill on the commissioned native language acoustic model, where we propose a data-driven method to make the seed lexicon which is used to train grapheme-to-phoneme model to predict mapping pronunciations for foreign language word in code-switching sentences. The core work of the data-driven technology in this paper consists of a phonetic decoding method and different selection methods. And for imbalanced word-level driving materials problem, we have an internal assistance inspiration that learning the good pronunciation rules in the words that possess sufficient materials using the grapheme-to-phoneme model to help the scarce. Our experiments show that the Mixed Error Rate in intra-sentential Chinese-English code-switching recognition reduced from 29.15\%, acquired on the pure Chinese recognizer, to 12.13\% by adding foreign language words' pronunciation through our data-driven approach, and finally get the best result 11.14\% with the combination of different selection methods and internal assistance tactic.

cs.SD↗

arXiv2022

Self-Adaptive Training: Bridging Supervised and Self-Supervised Learning

We propose self-adaptive training -- a unified training algorithm that dynamically calibrates and enhances training processes by model predictions without incurring an extra computational cost -- to advance both supervised and self-supervised learning of deep neural networks. We analyze the training dynamics of deep networks on training data that are corrupted by, e.g., random noise and adversarial examples. Our analysis shows that model predictions are able to magnify useful underlying information in data and this phenomenon occurs broadly even in the absence of any label information, highlighting that model predictions could substantially benefit the training processes: self-adaptive training improves the generalization of deep networks under noise and enhances the self-supervised representation learning. The analysis also sheds light on understanding deep learning, e.g., a potential explanation of the recently-discovered double-descent phenomenon in empirical risk minimization and the collapsing issue of the state-of-the-art self-supervised learning algorithms. Experiments on the CIFAR, STL, and ImageNet datasets verify the effectiveness of our approach in three applications: classification with label noise, selective classification, and linear evaluation. To facilitate future research, the code has been made publicly available at https://github.com/LayneH/self-adaptive-training.

cs.LG↗

arXiv2022

Multi-Objective Personalized Product Retrieval in Taobao Search

In large-scale e-commerce platforms like Taobao, it is a big challenge to retrieve products that satisfy users from billions of candidates. This has been a common concern of academia and industry. Recently, plenty of works in this domain have achieved significant improvements by enhancing embedding-based retrieval (EBR) methods, including the Multi-Grained Deep Semantic Product Retrieval (MGDSPR) model [16] in Taobao search engine. However, we find that MGDSPR still has problems of poor relevance and weak personalization compared to other retrieval methods in our online system, such as lexical matching and collaborative filtering. These problems promote us to further strengthen the capabilities of our EBR model in both relevance estimation and personalized retrieval. In this paper, we propose a novel Multi-Objective Personalized Product Retrieval (MOPPR) model with four hierarchical optimization objectives: relevance, exposure, click and purchase. We construct entire-space multi-positive samples to train MOPPR, rather than the single-positive samples for existing EBR models.We adopt a modified softmax loss for optimizing multiple objectives. Results of extensive offline and online experiments show that MOPPR outperforms the baseline MGDSPR on evaluation metrics of relevance estimation and personalized retrieval. MOPPR achieves 0.96% transaction and 1.29% GMV improvements in a 28-day online A/B test. Since the Double-11 shopping festival of 2021, MOPPR has been fully deployed in mobile Taobao search, replacing the previous MGDSPR. Finally, we discuss several advanced topics of our deeper explorations on multi-objective retrieval and ranking to contribute to the community.

cs.IR↗

arXiv2022

Equivalence of weak and viscosity solutions for the nonhomogeneous double phase equation

We establish the equivalence between weak and viscosity solutions to the nonhomogeneous double phase equation with lower-order term $$ -{\rm div}(|Du|^{p-2}Du+a(x)|Du|^{q-2}Du)=f(x,u,Du),\quad 1<p\le q<\infty, a(x)\ge0. $$ We find some appropriate hypotheses on the coefficient $a(x)$, the exponents $p, q$ and the nonlinear term $f$ to show that the viscosity solutions with {\em a priori} Lipschitz continuity are weak solutions of such equation by virtue of the $\inf$($\sup$)-convolution techniques. The reverse implication can be concluded through comparison principles. Moreover, we verify that the bounded viscosity solutions are exactly Lipschitz continuous, which is also of independent interest.

math.AP↗

arXiv2022

Expediting Large-Scale Vision Transformer for Dense Prediction without Fine-tuning

Vision transformers have recently achieved competitive results across various vision tasks but still suffer from heavy computation costs when processing a large number of tokens. Many advanced approaches have been developed to reduce the total number of tokens in large-scale vision transformers, especially for image classification tasks. Typically, they select a small group of essential tokens according to their relevance with the class token, then fine-tune the weights of the vision transformer. Such fine-tuning is less practical for dense prediction due to the much heavier computation and GPU memory cost than image classification. In this paper, we focus on a more challenging problem, i.e., accelerating large-scale vision transformers for dense prediction without any additional re-training or fine-tuning. In response to the fact that high-resolution representations are necessary for dense prediction, we present two non-parametric operators, a token clustering layer to decrease the number of tokens and a token reconstruction layer to increase the number of tokens. The following steps are performed to achieve this: (i) we use the token clustering layer to cluster the neighboring tokens together, resulting in low-resolution representations that maintain the spatial structures; (ii) we apply the following transformer layers only to these low-resolution representations or clustered tokens; and (iii) we use the token reconstruction layer to re-create the high-resolution representations from the refined low-resolution representations. The results obtained by our method are promising on five dense prediction tasks, including object detection, semantic segmentation, panoptic segmentation, instance segmentation, and depth estimation.

cs.CV↗

arXiv2022

Quantum variational learning for quantum error-correcting codes

Quantum error correction is believed to be a necessity for large-scale fault-tolerant quantum computation. In the past two decades, various constructions of quantum error-correcting codes (QECCs) have been developed, leading to many good code families. However, the majority of these codes are not suitable for near-term quantum devices. Here we present VarQEC, a noise-resilient variational quantum algorithm to search for quantum codes with a hardware-efficient encoding circuit. The cost functions are inspired by the most general and fundamental requirements of a QECC, the Knill-Laflamme conditions. Given the target noise channel (or the target code parameters) and the hardware connectivity graph, we optimize a shallow variational quantum circuit to prepare the basis states of an eligible code. In principle, VarQEC can find quantum codes for any error model, whether additive or non-additive, degenerate or non-degenerate, pure or impure. We have verified its effectiveness by (re)discovering some symmetric and asymmetric codes, e.g., $((n,2^{n-6},3))_2$ for $n$ from 7 to 14. We also found new $((6,2,3))_2$ and $((7,2,3))_2$ codes that are not equivalent to any stabilizer code, and extensive numerical evidence with VarQEC suggests that a $((7,3,3))_2$ code does not exist. Furthermore, we found many new channel-adaptive codes for error models involving nearest-neighbor correlated errors. Our work sheds new light on the understanding of QECC in general, which may also help to enhance near-term device performance with channel-adaptive error-correcting codes.

quant-ph↗

arXiv2022

Goal-Oriented Quantization: Analysis, Design, and Application to Resource Allocation

In this paper, the situation in which a receiver has to execute a task from a quantized version of the information source of interest is considered. The task is modeled by the minimization problem of a general goal function $f(x;g)$ for which the decision $x$ has to be taken from a quantized version of the parameters $g$. This problem is relevant in many applications e.g., for radio resource allocation (RA), high spectral efficiency communications, controlled systems, or data clustering in the smart grid. By resorting to high resolution (HR) analysis, it is shown how to design a quantizer that minimizes the gap between the minimum of $f$ (which would be reached by knowing $g$ perfectly) and what is effectively reached with a quantized $g$. The conducted formal analysis both provides quantization strategies in the HR regime and insights for the general regime and allows a practical algorithm to be designed. The analysis also allows one to provide some elements to the new and fundamental problem of the relationship between the goal function regularity properties and the hardness to quantize its parameters. The derived results are discussed and supported by a rich numerical performance analysis in which known RA goal functions are studied and allows one to exhibit very significant improvements by tailoring the quantization operation to the final task.

eess.SP↗

arXiv2022

Ta4CoSi: a Tantalum-rich superconductor with a honeycomb network structure

The crystal structure of solid-state materials with unique lattices is considered one of the main factors determining physical properties, including superconductivity. Materials with honeycomb lattices have inspired intense research interest for their novel properties. A previously unknown compound, Ta4CoSi, are discussed herein and its crystal structure and superconducting properties. It crystallizes in a CuAl2-type structure (space group P4/mcm, No. 124) with the lattice parameters a = b = 6.17755(7)Å and c = 4.98443(8) Å, featuring honeycomb networks of Ta-Ta in (110) plane. Superconductivity is observed below the critical temperature of 2.45 K, while lower and upper critical magnetic fields are 9.86 mT and 0.84 T, respectively. Experiments and theoretical calculations show that the honeycomb lattices significantly influence the superconducting properties. This material may thus provide a new platform for investigating the exotic superconductivity of honeycomb networks.

cond-mat.supr-con↗

arXiv2022

A general-purpose material property data extraction pipeline from large polymer corpora using Natural Language Processing

The ever-increasing number of materials science articles makes it hard to infer chemistry-structure-property relations from published literature. We used natural language processing (NLP) methods to automatically extract material property data from the abstracts of polymer literature. As a component of our pipeline, we trained MaterialsBERT, a language model, using 2.4 million materials science abstracts, which outperforms other baseline models in three out of five named entity recognition datasets when used as the encoder for text. Using this pipeline, we obtained ~300,000 material property records from ~130,000 abstracts in 60 hours. The extracted data was analyzed for a diverse range of applications such as fuel cells, supercapacitors, and polymer solar cells to recover non-trivial insights. The data extracted through our pipeline is made available through a web platform at https://polymerscholar.org which can be used to locate material property data recorded in abstracts conveniently. This work demonstrates the feasibility of an automatic pipeline that starts from published literature and ends with a complete set of extracted material property information.

cs.CL↗

arXiv2022

Context-Aware Query Rewriting for Improving Users' Search Experience on E-commerce Websites

E-commerce queries are often short and ambiguous. Consequently, query understanding often uses query rewriting to disambiguate user-input queries. While using e-commerce search tools, users tend to enter multiple searches, which we call context, before purchasing. These history searches contain contextual insights about users' true shopping intents. Therefore, modeling such contextual information is critical to a better query rewriting model. However, existing query rewriting models ignore users' history behaviors and consider only the instant search query, which is often a short string offering limited information about the true shopping intent. We propose an end-to-end context-aware query rewriting model to bridge this gap, which takes the search context into account. Specifically, our model builds a session graph using the history search queries and their contained words. We then employ a graph attention mechanism that models cross-query relations and computes contextual information of the session. The model subsequently calculates session representations by combining the contextual information with the instant search query using an aggregation network. The session representations are then decoded to generate rewritten queries. Empirically, we demonstrate the superiority of our method to state-of-the-art approaches under various metrics. On in-house data from an online shopping platform, by introducing contextual information, our model achieves 11.6% improvement under the MRR (Mean Reciprocal Rank) metric and 20.1% improvement under the HIT@16 metric (a hit rate metric), in comparison with the best baseline method (Transformer-based model).

cs.IR↗

arXiv2022

FlowFormer: A Transformer Architecture for Optical Flow

We introduce optical Flow transFormer, dubbed as FlowFormer, a transformer-based neural network architecture for learning optical flow. FlowFormer tokenizes the 4D cost volume built from an image pair, encodes the cost tokens into a cost memory with alternate-group transformer (AGT) layers in a novel latent space, and decodes the cost memory via a recurrent transformer decoder with dynamic positional cost queries. On the Sintel benchmark, FlowFormer achieves 1.159 and 2.088 average end-point-error (AEPE) on the clean and final pass, a 16.5% and 15.5% error reduction from the best published result (1.388 and 2.47). Besides, FlowFormer also achieves strong generalization performance. Without being trained on Sintel, FlowFormer achieves 1.01 AEPE on the clean pass of Sintel training set, outperforming the best published result (1.29) by 21.7%.

cs.CV↗

arXiv2022

Performance Analysis of Reconfigurable Intelligent Surface Assisted Two-Way NOMA Networks

This paper investigates the performance of reconfigurable intelligent surface assisted two-way non-orthogonal multiple access (RIS-TW-NOMA) networks, where a pair of users exchange their information through a RIS. The influence of imperfect successive interference cancellation on RIS-TW-NOMA is taken into account. To evaluate the potential performance of RIS-TW-NOMA, we derive the exact and asymptotic expressions of outage probability and ergodic rate for a pair of users. Based on the analytical results, the diversity orders and high signal-to-noise ratio (SNR) slopes are obtained in the high SNR regime, which are closely related to the number of RIS elements. Additionally, we analyze the system throughput and energy efficiency of RIS-TW-NOMA networks in both delay-limited and delay-tolerant transmission modes. Numerical results indicate that: 1) The outage behaviors and ergodic rate of RIS-TW-NOMA are superior to that of RIS-TW-OMA and two-way relay OMA (TWR-OMA); 2) As the number of RIS elements increases, the RIS-TW-NOMA networks are capable of achieving the enhanced outage performance; and 3) By comparing with RIS-TW-OMA and TWR-OMA networks, the energy efficiency and system throughput of RIS-TW-NOMA has obvious advantages.

cs.IT↗

arXiv2022

Streaming End-to-End Multilingual Speech Recognition with Joint Language Identification

Language identification is critical for many downstream tasks in automatic speech recognition (ASR), and is beneficial to integrate into multilingual end-to-end ASR as an additional task. In this paper, we propose to modify the structure of the cascaded-encoder-based recurrent neural network transducer (RNN-T) model by integrating a per-frame language identifier (LID) predictor. RNN-T with cascaded encoders can achieve streaming ASR with low latency using first-pass decoding with no right-context, and achieve lower word error rates (WERs) using second-pass decoding with longer right-context. By leveraging such differences in the right-contexts and a streaming implementation of statistics pooling, the proposed method can achieve accurate streaming LID prediction with little extra test-time cost. Experimental results on a voice search dataset with 9 language locales shows that the proposed method achieves an average of 96.2% LID prediction accuracy and the same second-pass WER as that obtained by including oracle LID in the input.

eess.AS↗

arXiv2022

A Q-band line survey towards Orion KL using the Tianma radio telescope

We have conducted a line survey towards Orion KL using the Q-band receiver of Tianma 65 m radio telescope (TMRT), covering 34.8--50 GHz with a velocity resolution between 0.79 km s$^{-1}$ and 0.55 km s$^{-1}$ respectively. The observations reach a sensitivity on the level of 1-8 mK, proving that the TMRT is sensitive for conducting deep line surveys. In total, 597 Gaussian features are extracted. Among them, 177 radio recombination lines (RRLs) are identified, including 126, 40 and 11 RRLs of hydrogen, helium and carbon, with a maximum $Δn$ of 16, 7, and 3, respectively. The carbon RRLs are confirmed to originate from photodissociation regions with a $V_{\rm LSR}\sim$9 km s$^{-1}$. In addition, 371 molecular transitions of 53 molecular species are identified. Twenty-one molecular species of this survey were not firmly detected in the Q band by Rizzo et al. (2017), including species such as H$_2$CS, HCOOH, C$_2$H$_5$OH, H$_2^{13}$CO, H$_2$CCO, CH$_3$CHO, CH$_2$OCH$_2$, HCN $v_2=1$, and CH$_3$OCHO $v_t=1$. In particular, the vibrationally excited states of ethyl cyanide (C$_2$H$_5$CN $v$13/$v$21) are for the first time firmly detected in the Q band. NH$_3$ (15,15) and (16,16) are identified, and they are so far the highest transitions of the NH$_3$ inversion lines detected towards Orion KL. All the identified lines can be reproduced by a radiative transfer model.

astro-ph.GA↗

arXiv2022

SaleNet: A low-power end-to-end CNN accelerator for sustained attention level evaluation using EEG

This paper proposes SaleNet - an end-to-end convolutional neural network (CNN) for sustained attention level evaluation using prefrontal electroencephalogram (EEG). A bias-driven pruning method is proposed together with group convolution, global average pooling (GAP), near-zero pruning, weight clustering and quantization for the model compression, achieving a total compression ratio of 183.11x. The compressed SaleNet obtains a state-of-the-art subject-independent sustained attention level classification accuracy of 84.2% on the recorded 6-subject EEG database in this work. The SaleNet is implemented on a Artix-7 FPGA with a competitive power consumption of 0.11 W and an energy-efficiency of 8.19 GOps/W.

cs.AR↗