arXiv subjects

Peng Xu

Publications and source records attributed to Peng Xu.

At least 343 records · Page 19Linked to original sources

arXiv2020

TRANS-BLSTM: Transformer with Bidirectional LSTM for Language Understanding

Bidirectional Encoder Representations from Transformers (BERT) has recently achieved state-of-the-art performance on a broad range of NLP tasks including sentence classification, machine translation, and question answering. The BERT model architecture is derived primarily from the transformer. Prior to the transformer era, bidirectional Long Short-Term Memory (BLSTM) has been the dominant modeling architecture for neural machine translation and question answering. In this paper, we investigate how these two modeling techniques can be combined to create a more powerful model architecture. We propose a new architecture denoted as Transformer with BLSTM (TRANS-BLSTM) which has a BLSTM layer integrated to each transformer block, leading to a joint modeling framework for transformer and BLSTM. We show that TRANS-BLSTM models consistently lead to improvements in accuracy compared to BERT baselines in GLUE and SQuAD 1.1 experiments. Our TRANS-BLSTM model obtains an F1 score of 94.01% on the SQuAD 1.1 development dataset, which is comparable to the state-of-the-art result.

cs.CL↗

arXiv2020

Learning Fast Adaptation on Cross-Accented Speech Recognition

Local dialects influence people to pronounce words of the same language differently from each other. The great variability and complex characteristics of accents creates a major challenge for training a robust and accent-agnostic automatic speech recognition (ASR) system. In this paper, we introduce a cross-accented English speech recognition task as a benchmark for measuring the ability of the model to adapt to unseen accents using the existing CommonVoice corpus. We also propose an accent-agnostic approach that extends the model-agnostic meta-learning (MAML) algorithm for fast adaptation to unseen accents. Our approach significantly outperforms joint training in both zero-shot, few-shot, and all-shot in the mixed-region and cross-region settings in terms of word error rate.

eess.AS↗

arXiv2020

Hierarchical emotion-recognition framework based on discriminative brain neural network topology and ensemble co-decision strategy

Brain neural networks characterize various information propagation patterns for different emotional states. However, the statistical features based on traditional graph theory may ignore the spacial network difference. To reveal these inherent spatial features and increase the stability of emotional recognition, we proposed a hierarchical framework that can perform the multiple emotion recognitions with the multiple emotion-related spatial network topology patterns (MESNP) by combining a supervised learning with ensemble co-decision strategy. To evaluate the performance of our proposed MESNP approach, we conduct both off-line and simulated on-line experiments with two public datasets i.e., MAHNOB and DEAP. The experiment results demonstrated that MESNP can significantly enhance the classification performance for the multiple emotions. The highest accuracies of off-line experiments for MAHNOB-HCI and DEAP achieved 99.93% (3 classes) and 83.66% (4 classes), respectively. For simulated on-line experiments, we also obtained the best classification accuracies with 100% (3 classes) for MAHNOB and 99.22% (4 classes) for DEAP by proposed MESNP. These results further proved the efficiency of MESNP for structured feature extraction in mult-classification emotional task.

q-bio.NC↗

arXiv2020

Better Long-Range Dependency By Bootstrapping A Mutual Information Regularizer

In this work, we develop a novel regularizer to improve the learning of long-range dependency of sequence data. Applied on language modelling, our regularizer expresses the inductive bias that sequence variables should have high mutual information even though the model might not see abundant observations for complex long-range dependency. We show how the `next sentence prediction (classification)' heuristic can be derived in a principled way from our mutual information estimation framework, and be further extended to maximize the mutual information of sequence variables. The proposed approach not only is effective at increasing the mutual information of segments under the learned model but more importantly, leads to a higher likelihood on holdout data, and improved generation quality. Code is released at https://github.com/BorealisAI/BMI.

cs.LG↗

arXiv2020

Fine-Grained Instance-Level Sketch-Based Video Retrieval

Existing sketch-analysis work studies sketches depicting static objects or scenes. In this work, we propose a novel cross-modal retrieval problem of fine-grained instance-level sketch-based video retrieval (FG-SBVR), where a sketch sequence is used as a query to retrieve a specific target video instance. Compared with sketch-based still image retrieval, and coarse-grained category-level video retrieval, this is more challenging as both visual appearance and motion need to be simultaneously matched at a fine-grained level. We contribute the first FG-SBVR dataset with rich annotations. We then introduce a novel multi-stream multi-modality deep network to perform FG-SBVR under both strong and weakly supervised settings. The key component of the network is a relation module, designed to prevent model over-fitting given scarce training data. We show that this model significantly outperforms a number of existing state-of-the-art models designed for video analysis.

cs.CV↗

arXiv2020

Scalable quantum computing with qudits on a graph

We show a significant reduction of the number of quantum operations and the improvement of the circuit depth for the realization of the Toffoli gate by using qudits. This is done by establishing a general relation between the dimensionality of qudits and their topology of connections for a scalable multi-qudit processor, where higher qudit levels are used for substituting ancillas. The suggested model is of importance for the realization of quantum algorithms and as a method of quantum error correction codes for single-qubit operations.

quant-ph↗

arXiv2020

Deep Self-Supervised Representation Learning for Free-Hand Sketch

In this paper, we tackle for the first time, the problem of self-supervised representation learning for free-hand sketches. This importantly addresses a common problem faced by the sketch community -- that annotated supervisory data are difficult to obtain. This problem is very challenging in that sketches are highly abstract and subject to different drawing styles, making existing solutions tailored for photos unsuitable. Key for the success of our self-supervised learning paradigm lies with our sketch-specific designs: (i) we propose a set of pretext tasks specifically designed for sketches that mimic different drawing styles, and (ii) we further exploit the use of a textual convolution network (TCN) in a dual-branch architecture for sketch feature learning, as means to accommodate the sequential stroke nature of sketches. We demonstrate the superiority of our sketch-specific designs through two sketch-related applications (retrieval and recognition) on a million-scale sketch dataset, and show that the proposed approach outperforms the state-of-the-art unsupervised representation learning methods, and significantly narrows the performance gap between with supervised representation learning.

cs.CV↗

arXiv2020

EEG-based Brain-Computer Interfaces (BCIs): A Survey of Recent Studies on Signal Sensing Technologies and Computational Intelligence Approaches and their Applications

Brain-Computer Interface (BCI) is a powerful communication tool between users and systems, which enhances the capability of the human brain in communicating and interacting with the environment directly. Advances in neuroscience and computer science in the past decades have led to exciting developments in BCI, thereby making BCI a top interdisciplinary research area in computational neuroscience and intelligence. Recent technological advances such as wearable sensing devices, real-time data streaming, machine learning, and deep learning approaches have increased interest in electroencephalographic (EEG) based BCI for translational and healthcare applications. Many people benefit from EEG-based BCIs, which facilitate continuous monitoring of fluctuations in cognitive states under monotonous tasks in the workplace or at home. In this study, we survey the recent literature of EEG signal sensing technologies and computational intelligence approaches in BCI applications, compensated for the gaps in the systematic summary of the past five years (2015-2019). In specific, we first review the current status of BCI and its significant obstacles. Then, we present advanced signal sensing and enhancement technologies to collect and clean EEG signals, respectively. Furthermore, we demonstrate state-of-art computational intelligence techniques, including interpretable fuzzy models, transfer learning, deep learning, and combinations, to monitor, maintain, or track human cognitive states and operating performance in prevalent applications. Finally, we deliver a couple of innovative BCI-inspired healthcare applications and discuss some future research directions in EEG-based BCIs.

eess.SP↗

arXiv2020

Quantum-enhanced interferometry with asymmetric beam splitters

In this paper, we investigate the phase sensitivities in two-path optical interferometry with asymmetric beam splitters. Here, we present the optimal conditions for the transmission ratio and the phase of the beam splitter to gain the highest sensitivities for a general class of non-classical states with parity symmetry. Additionally, we address the controversial question of whether the scheme with a combination of coherent state and photon-added or photon-subtracted squeezed vacuum state is better or worse than the most celebrated one using a combination of coherent state and squeezed vacuum state.

quant-ph↗

arXiv2019

SPIN: A High Speed, High Resolution Vision Dataset for Tracking and Action Recognition in Ping Pong

We introduce a new high resolution, high frame rate stereo video dataset, which we call SPIN, for tracking and action recognition in the game of ping pong. The corpus consists of ping pong play with three main annotation streams that can be used to learn tracking and action recognition models -- tracking of the ping pong ball and poses of humans in the videos and the spin of the ball being hit by humans. The training corpus consists of 53 hours of data with labels derived from previous models in a semi-supervised method. The testing corpus contains 1 hour of data with the same information, except that crowd compute was used to obtain human annotations of the ball position, from which ball spin has been derived. Along with the dataset we introduce several baseline models that were trained on this data. The models were specifically chosen to be able to perform inference at the same rate as the images are generated -- specifically 150 fps. We explore the advantages of multi-task training on this data, and also show interesting properties of ping pong ball trajectories that are derived from our observational data, rather than from prior physics models. To our knowledge this is the first large scale dataset of ping pong; we offer it to the community as a rich dataset that can be used for a large variety of machine learning and vision tasks such as tracking, pose estimation, semi-supervised and unsupervised learning and generative modeling.

cs.CV↗

arXiv2019

Doppler-insensitive two-qubit controlled-PHASE gate protocol with dual-pulse off-resonant modulated driving for neutral atoms

For neutral atom qubits, the residual thermal motion of the cold atoms constitutes a major challenge that limits the accessible two-qubit gate fidelity. Recently, an interesting type of two-qubit controlled-PHASE quantum gate protocol has been introduced for neutral atom qubit platform, which relies upon off-resonant modulated driving and Rydberg blockade effect. Building upon this progress, we have further developed an upgrade in the form of dual-pulse off-resonant modulated driving. Besides the inherent advantages of avoiding shelving population in Rydberg levels, not necessarily requiring individual site addressing, not sensitive to the exact value of blockade shift while suppressing population leakage error and rotation error, the major new feature of this protocol is Doppler-insensitive. In principle, the gate fidelity remains reasonably high over a relatively significant velocity range of the qubit atoms. Moreover, we anticipate that this protocol will inspire future improvements in quantum gate protocols for other types of qubit platforms, and its strategies may find applications in the area of quantum optimal control.

quant-ph↗

arXiv2019

Preparation of a Heteronuclear Two-atom System in the 3D Motional Ground State in an Optical Tweezer

We report the realization of a heteronuclear two-atom of $^{87}$Rb-$^{85}$Rb in the ground state of an optical tweezer (OT). Starting by trapping two different isotopic single atoms, a $^{87}$Rb and a $^{85}$Rb in two strongly focused and linearly polarized OT with 4 $μ$m apart, we perform simultaneously three dimensional Raman sideband cooling for both atoms and the obtained 3D ground state probabilities of $^{87}$Rb and $^{85}$Rb are 0.91(5) and 0.91(10) respectively. There is no obvious crosstalk observed during the cooling process. We then merge them into one tweezer via a species-dependent transport, where the species-dependent potentials are made by changing the polarization of the OTs for each species from linear polarization to the desired circular polarization. The measurable increment of vibrational quantum due to merging is $0.013(1)$ for the axial dimension. This two-atom system can be used to investigate cold collisional physics, to form quantum logic gates, and to build a single heteronuclear molecule. It can also be scaled up to few-atom regime and extended to other atomic species and molecules, and thus to ultracold chemistry.

physics.atom-ph↗

arXiv2019

Integration of the Berry curvature on a qubit state manifold by coupling to a quantum meter system

We present a scheme that allows integration of the Berry curvature and thus determination of the Chern number of a qubit eigenstate manifold. Our proposal continuously couples the qubit with a meter system while it explores a quasi-adiabatic path in the manifold. The accumulated change of one of the meter observables then provides an estimate of the Chern number. By varying the initial state of the meter, we explore the delicate interplay between the measurement precision and the disturbance of the qubit. A simple argument yields a correction factor that allows estimation of the Chern number, even when the qubit is significantly disturbed during the probing. The Chern number arises from the geometric phase accumulated during the exploration, while we observe the dynamic phase to produce a broadening of the meter wave function. We show that a protocol, relying on three subsequent explorations, allows cancellation of the dynamic phase while the geometric phase is retained.

quant-ph↗

arXiv2019

Attention-Informed Mixed-Language Training for Zero-shot Cross-lingual Task-oriented Dialogue Systems

Recently, data-driven task-oriented dialogue systems have achieved promising performance in English. However, developing dialogue systems that support low-resource languages remains a long-standing challenge due to the absence of high-quality data. In order to circumvent the expensive and time-consuming data collection, we introduce Attention-Informed Mixed-Language Training (MLT), a novel zero-shot adaptation method for cross-lingual task-oriented dialogue systems. It leverages very few task-related parallel word pairs to generate code-switching sentences for learning the inter-lingual semantics across languages. Instead of manually selecting the word pairs, we propose to extract source words based on the scores computed by the attention layer of a trained English task-related model and then generate word pairs using existing bilingual dictionaries. Furthermore, intensive experiments with different cross-lingual embeddings demonstrate the effectiveness of our approach. Finally, with very few word pairs, our model achieves significant zero-shot adaptation performance improvements in both cross-lingual dialogue state tracking and natural language understanding (i.e., intent detection and slot filling) tasks compared to the current state-of-the-art approaches, which utilize a much larger amount of bilingual data.

cs.CL↗

arXiv2019

Optimal Number of Clusters by Measuring Similarity among Topographies for Spatio-temporal ERP Analysis

Averaging amplitudes over consecutive time samples within a time-window is widely used to calculate the amplitude of an event-related potential (ERP) for cognitive neuroscience. Objective determination of the time-window is critical for determining the ERP component. Clustering on the spatio-temporal ERP data can obtain the time-window in which the consecutive time samples topographies are expected to be highly similar in practice. However, there exists a challenging problem of determining an optimal number of clusters. Here, we develop a novel methodology to obtain the optimal number of clusters using consensus clustering on the spatio-temporal ERP data. Various clustering methods, namely, K-means, hierarchical clustering, fuzzy C-means, self-organizing map, and diffusion maps spectral clustering are combined in an ensemble clustering manner to find the most reliable clusters. When a range of numbers of clusters is applied on the spatio-temporal ERP dataset, the optimal number of clusters should correspond to the cluster of interest within which the average of correlation coefficients between topographies of every two-time sample in the time-window is the maximum for an ERP of interest. In our method, we consider fewer cluster maps for analyzing an optimal number of clusters for isolating the components of interest in the spatio-temporal ERP. The statistical comparison demonstrates that the present method outperforms other conventional approaches. This finding would be practically useful for discovering the optimal clustering in spatio-temporal ERP, especially when the cognitive knowledge about time-window is not clearly defined.

q-bio.NC↗

arXiv2019

Zero-shot Cross-lingual Dialogue Systems with Transferable Latent Variables

Despite the surging demands for multilingual task-oriented dialog systems (e.g., Alexa, Google Home), there has been less research done in multilingual or cross-lingual scenarios. Hence, we propose a zero-shot adaptation of task-oriented dialogue system to low-resource languages. To tackle this challenge, we first use a set of very few parallel word pairs to refine the aligned cross-lingual word-level representations. We then employ a latent variable model to cope with the variance of similar sentences across different languages, which is induced by imperfect cross-lingual alignments and inherent differences in languages. Finally, the experimental results show that even though we utilize much less external resources, our model achieves better adaptation performance for natural language understanding task (i.e., the intent detection and slot filling) compared to the current state-of-the-art model in the zero-shot scenario.

cs.CL↗

arXiv2019

Two-qubit controlled-PHASE Rydberg blockade gate protocol for neutral atoms via off-resonant modulated driving within a single pulse

Neutral atom array serves as an ideal platform to study the quantum logic gates, where intense efforts have been devoted to improve the two-qubit gate fidelity. We report our recent findings in constructing a different type of two-qubit controlled-PHASE quantum gate protocol with neutral atoms enabled by Rydberg blockade, which aims at both robustness and high-fidelity. It relies upon modulated driving pulse with specially tailored smooth waveform to gain appropriate phase accumulations for quantum gates. The major features include finishing gate operation within a single pulse, not necessarily requiring individual site addressing, not sensitive to the exact value of blockade shift while suppressing population leakage error and rotation error. We anticipate its fidelity to be reasonably high under realistic considerations for errors such as atomic motion, laser power fluctuation, power imbalance, spontaneous emission and so on. Moreover, we hope that such type of protocol may inspire future improvements in quantum gate designs for other categories of qubit platforms and new applications in other areas of quantum optimal control.

quant-ph↗

arXiv2019

Boosting Image Recognition with Non-differentiable Constraints

In this paper, we study the problem of image recognition with non-differentiable constraints. A lot of real-life recognition applications require a rich output structure with deterministic constraints that are discrete or modeled by a non-differentiable function. A prime example is recognizing digit sequences, which are restricted by such rules (e.g., \textit{container code detection}, \textit{social insurance number recognition}, etc.). We investigate the usefulness of adding non-differentiable constraints in learning for the task of digit sequence recognition. Toward this goal, we synthesize six different datasets from MNIST and Cropped SVHN, with three discrete rules inspired by real-life protocols. To deal with the non-differentiability of these rules, we propose a reinforcement learning approach based on the policy gradient method. We find that incorporating this rule-based reinforcement can effectively increase the accuracy for all datasets and provide a good inductive bias which improves the model even with limited data. On one of the datasets, MNIST\_Rule2, models trained with rule-based reinforcement increase the accuracy by 4.7\% for 2000 samples and 23.6\% for 500 samples. We further test our model against synthesized adversarial examples, e.g., blocking out digits, and observe that adding our rule-based reinforcement increases the model robustness with a relatively smaller performance drop.

cs.CV↗