arXiv subjects

Nan Yang

Publications and source records attributed to Nan Yang.

At least 181 records · Page 10Linked to original sources

arXiv2019

Unified Language Model Pre-training for Natural Language Understanding and Generation

This paper presents a new Unified pre-trained Language Model (UniLM) that can be fine-tuned for both natural language understanding and generation tasks. The model is pre-trained using three types of language modeling tasks: unidirectional, bidirectional, and sequence-to-sequence prediction. The unified modeling is achieved by employing a shared Transformer network and utilizing specific self-attention masks to control what context the prediction conditions on. UniLM compares favorably with BERT on the GLUE benchmark, and the SQuAD 2.0 and CoQA question answering tasks. Moreover, UniLM achieves new state-of-the-art results on five natural language generation datasets, including improving the CNN/DailyMail abstractive summarization ROUGE-L to 40.51 (2.04 absolute improvement), the Gigaword abstractive summarization ROUGE-L to 35.75 (0.86 absolute improvement), the CoQA generative question answering F1 score to 82.5 (37.1 absolute improvement), the SQuAD question generation BLEU-4 to 22.12 (3.75 absolute improvement), and the DSTC7 document-grounded dialog response generation NIST-4 to 2.67 (human performance is 2.65). The code and pre-trained models are available at https://github.com/microsoft/unilm.

cs.CL↗

arXiv2019

Multi-Frame GAN: Image Enhancement for Stereo Visual Odometry in Low Light

We propose the concept of a multi-frame GAN (MFGAN) and demonstrate its potential as an image sequence enhancement for stereo visual odometry in low light conditions. We base our method on an invertible adversarial network to transfer the beneficial features of brightly illuminated scenes to the sequence in poor illumination without costly paired datasets. In order to preserve the coherent geometric cues for the translated sequence, we present a novel network architecture as well as a novel loss term combining temporal and stereo consistencies based on optical flow estimation. We demonstrate that the enhanced sequences improve the performance of state-of-the-art feature-based and direct stereo visual odometry methods on both synthetic and real datasets in challenging illumination. We also show that MFGAN outperforms other state-of-the-art image enhancement and style transfer methods by a large margin in terms of visual odometry.

cs.RO↗

arXiv2019

Expected Density of Cooperative Bacteria in a 2D Quorum Sensing Based Molecular Communication System

The exchange of small molecular signals within microbial populations is generally referred to as quorum sensing (QS). QS is ubiquitous in nature and enables microorganisms to respond to fluctuations in living environments by working together. In this study, a QS-based molecular communication system within a microbial population in a two-dimensional (2D) environment is analytically modeled. Microorganisms are randomly distributed on a 2D circle where each one releases molecules at random times. The number of molecules observed at each randomly-distributed bacterium is first derived by characterizing the diffusion and degradation of signaling molecules within the population. Using the derived result and some approximation, the expected density of cooperative bacteria is derived. Our model captures the basic features of QS. The analytical results for noisy signal propagation agree with simulation results where the Brownian motion of molecules is simulated by a particle-based method. Therefore, we anticipate that our model can be used to predict the density of cooperative bacteria in a variety of QS-coordinated activities, e.g., biofilm formation and antibiotic resistance.

q-bio.CB↗

arXiv2019

One-Way URLLC with Truncated Channel Inversion Power Control

In this work, we consider one-way ultra-reliable and low-latency communication (URLLC), where only the transmission in one direction requires URLLC and the transmission in the opposite direction does not. In order to meet the low-latency requirement of the one-way URLLC, we propose to use a truncated channel inversion power control (CIPC) to eliminate the requirement and the associated overhead of the training-based channel estimation at the receiver, while utilizing the multi-antenna technique at the transmitter to enhance the communication reliability. We first derive the transmission outage probability achieved by the truncated CIPC by considering the impact of a finite blocklength and a maximum transmit power constraint. Then, we determine the optimal constant power of the received signals in the truncated CIPC, which minimizes the transmission outage probability. Our examination shows that the proposed truncated CIPC is an effective means to achieve the one-way URLLC, where the tradeoff among reliability, latency, and required resources (e.g., the required number of transmit antennas, or the required maximum transmit power) is revealed.

cs.IT↗

arXiv2019

Non-Locality and Zero-Knowledge MIPs

The foundation of zero-knowledge is the simulator: a weak machine capable of pretending to be a weak verifier talking with all-powerful provers. To achieve this, simulators need some kind of advantage such as the knowledge of a trapdoor. In existing zero-knowledge multi-prover protocols, this advantage is essentially signalling, something that the provers are explicitly forbidden to do. In most cases, this advantage is stronger than necessary as it is possible to define a sense in which simulators need much less to simulate. We define a framework in which we can quantify the simulators' non-local advantage and exhibit examples of zero-knowledge protocols that are sound against local or entangled provers but that are not sound against no-signalling provers precisely because the no-signalling simulation strategy can be adopted by malicious provers.

quant-ph↗

arXiv2019

Escape, bound and capture geodesics in local static coordinates in Schwarzschild spacetime

The classical geodesics of timelike particles in Schwarzschild spacetime is analyzed according to the particle starting radius $r$, velocity $v$ and angle $α$ against the radial outward direction in the reference system of an local static observer. The region of escape, bound and capture orbits in the parameter space of $(r,~v,~α)$ are solved using the three cases of the effective potential. It is found that generally for radius smaller than $4M$ or velocity larger than $c/\sqrt{2}$ there will be no bound orbits. While for fixed radius larger than $4M$ (or velocity smaller than $c/\sqrt{2}$), as velocity (or radius) increase from zero (or $2M$), the particle is always captured until a critical value $v_{\mathrm{crit1}}$ (or $r_{\mathrm{crit1}}$) when the bound orbit start to appear around $α=π/2$ between a double-napped cone structure. As the velocity (or radius) increases to another critical value $v_{\mathrm{crit2}}$ (or $r_{\mathrm{crit2}}$) then the bound directions and escape directions in the outward cone become escape directions, leaving only the inward cone separating the capture and bound directions. The angle of this cone will increase to its asymptotic value as velocity (or radius) increases to its asymptotic value. The implication of these results in shadow of black holes formed by massive particles, in black hole accretion and in spacecraft navigation is briefly discussed.

gr-qc↗

arXiv2019

Molecular Information Delivery in Porous Media

Information delivery via molecular signals is abundant in nature and potentially useful for industry sensing. Many propagation channels (e.g., tissue membranes and catalyst beds) contain porous medium materials and the impact this has on communication performance is not well understood. Here, communication through realistic porous channels is investigated for the first time via statistical breakthrough curves. Assuming that the number of arrived molecules can be approximated as a Gaussian random variable and using fully resolved computational fluid dynamics results for the breakthrough curves, the numerical results for the throughput, mutual information, error probability, and information diversity gain are presented. Using these numerical results, the unique characteristics of the porous medium channel are revealed.

cs.IT↗

arXiv2019

Non-Locality in Interactive Proofs

In multi-prover interactive proofs (MIPs), the verifier is usually non-adaptive. This stems from an implicit problem which we call ``contamination'' by the verifier. We make explicit the verifier contamination problem, and identify a solution by constructing a generalization of the MIP model. This new model quantifies non-locality as a new dimension in the characterization of MIPs. A new property of zero-knowledge emerges naturally as a result by also quantifying the non-locality of the simulator.

quant-ph↗

arXiv2019

Synaptic Learning and Memory Functions Achieved in Self-rectifying BFO Memristor under Extreme Environmental Temperature

Memristors have been intensively studied in recent years as promising building blocks for next-generation non-volatile memory, artificial neural networks and brain-inspired computing systems. Even though the environment adaptability of memristor has been required in many application fields, it has been rarely reported due to the underlying mechanism could become invalid especially at an elevated temperature. Here, we focus on achieving synaptic learning and memory functions in BiFeO3 memristor in a wide range of temperature. We have proved the ferroelectricity of BFO films at a record-high temperature of 500 °C by piezoresponse force microscopy (PFM) measurement. Due to the robust ferroelectricity of BFO thin film, an analog-like resistance switching behavior has been clearly found in a wide range of temperature, which is attributed to the reversal of ferroelectric polarization. Various synaptic functions including long-term potentiation (LTP), depression (LTD), consecutive potentiation/depression (P/D) and spike-timing dependent plasticity (STDP) have been realized from -170 to 300 °C, illustrating their potential for electronic applications even under extreme environmental temperature.

cond-mat.mtrl-sci↗

arXiv2019

Using Game Theory for Real-Time Behavioural Dynamics in Microscopic Populations with Noisy Signalling

This paper introduces the application of game theory to understand noisy real-time signalling and the resulting behavioural dynamics in microscopic populations such as bacteria and other cells. It presents a bridge between the fields of molecular communication and microscopic game theory. Molecular communication uses conventional communication engineering theory and techniques to study and design systems that use chemical molecules as information carriers. Microscopic game theory models interactions within and between populations of cells and microorganisms. Integrating these two fields provides unique opportunities to understand and control microscopic populations that have imperfect signal propagation. Two examples, namely bacteria quorum sensing and tumour cell signalling, are presented with potential games to demonstrate the application of this approach. Finally, a case study of bacteria resource sharing demonstrates how noisy signalling can alter the distribution of behaviour.

q-bio.CB↗

arXiv2019

Verifier Non-Locality in Interactive Proofs

In multi-prover interactive proofs, the verifier interrogates the provers and attempts to steal their knowledge. Other than that, the verifier's role has not been studied. We have discovered that the verifier plays a much more important role than previously thought. Simply put, the verifier has the capability of providing non-local resources for the provers intrinsically. Existing MIPs' proofs of soundness implicitly depend on the fact that the verifier is not a non-local resource provider. The verifier's non-locality is a new unused tool and liability for protocol design and analysis.

quant-ph↗

arXiv2018

A Novel A Priori Simulation Algorithm for Absorbing Receivers in Diffusion-Based Molecular Communication Systems

A novel a priori Monte Carlo (APMC) algorithm is proposed to accurately simulate the molecules absorbed at spherical receiver(s) with low computational complexity in diffusion-based molecular communication (MC) systems. It is demonstrated that the APMC algorithm achieves high simulation efficiency since by using this algorithm, the fraction of molecules absorbed for a relatively large time step length precisely matches the analytical result. Therefore, the APMC algorithm overcomes the shortcoming of the existing refined Monte Carlo (RMC) algorithm which enables accurate simulation for a relatively small time step length only. Moreover, for the RMC algorithm, an expression is proposed to quickly predict the simulation accuracy as a function of the time step length and system parameters, which facilitates the choice of simulation time step for a given system. Furthermore, a rejection threshold is proposed for both the RMC and APMC algorithms to significantly save computational complexity while causing an extremely small loss in accuracy.

cs.ET↗

arXiv2018

Symbol-by-Symbol Maximum Likelihood Detection for Cooperative Molecular Communication

In this paper, symbol-by-symbol maximum likelihood (ML) detection is proposed for a cooperative diffusion-based molecular communication (MC) system. In this system, the transmitter (TX) sends a common information symbol to multiple receivers (RXs) and a fusion center (FC) chooses the TX symbol that is more likely, given the likelihood of its observations from all RXs. The transmission of a sequence of binary symbols and the resultant intersymbol interference are considered in the cooperative MC system. Three ML detection variants are proposed according to different RX behaviors and different knowledge at the FC. The system error probabilities for two ML detector variants are derived, one of which is in closed form. The optimal molecule allocation among RXs to minimize the system error probability of one variant is determined by solving a joint optimization problem. Also for this variant, the equal distribution of molecules among two symmetric RXs is analytically shown to achieve the local minimal error probability. Numerical and simulation results show that the ML detection variants provide lower bounds on the error performance of simpler, non-ML cooperative variants and demonstrate that these simpler cooperative variants have error performance comparable to ML detectors.

cs.IT↗

arXiv2018

Read + Verify: Machine Reading Comprehension with Unanswerable Questions

Machine reading comprehension with unanswerable questions aims to abstain from answering when no answer can be inferred. In addition to extract answers, previous works usually predict an additional "no-answer" probability to detect unanswerable cases. However, they fail to validate the answerability of the question by verifying the legitimacy of the predicted answer. To address this problem, we propose a novel read-then-verify system, which not only utilizes a neural reader to extract candidate answers and produce no-answer probabilities, but also leverages an answer verifier to decide whether the predicted answer is entailed by the input snippets. Moreover, we introduce two auxiliary losses to help the reader better handle answer extraction as well as no-answer detection, and investigate three different architectures for the answer verifier. Our experiments on the SQuAD 2.0 dataset show that our system achieves a score of 74.2 F1 on the test set, achieving state-of-the-art results at the time of submission (Aug. 28th, 2018).

cs.CL↗

arXiv2018

A Disease Diagnosis and Treatment Recommendation System Based on Big Data Mining and Cloud Computing

It is crucial to provide compatible treatment schemes for a disease according to various symptoms at different stages. However, most classification methods might be ineffective in accurately classifying a disease that holds the characteristics of multiple treatment stages, various symptoms, and multi-pathogenesis. Moreover, there are limited exchanges and cooperative actions in disease diagnoses and treatments between different departments and hospitals. Thus, when new diseases occur with atypical symptoms, inexperienced doctors might have difficulty in identifying them promptly and accurately. Therefore, to maximize the utilization of the advanced medical technology of developed hospitals and the rich medical knowledge of experienced doctors, a Disease Diagnosis and Treatment Recommendation System (DDTRS) is proposed in this paper. First, to effectively identify disease symptoms more accurately, a Density-Peaked Clustering Analysis (DPCA) algorithm is introduced for disease-symptom clustering. In addition, association analyses on Disease-Diagnosis (D-D) rules and Disease-Treatment (D-T) rules are conducted by the Apriori algorithm separately. The appropriate diagnosis and treatment schemes are recommended for patients and inexperienced doctors, even if they are in a limited therapeutic environment. Moreover, to reach the goals of high performance and low latency response, we implement a parallel solution for DDTRS using the Apache Spark cloud platform. Extensive experimental results demonstrate that the proposed DDTRS realizes disease-symptom clustering effectively and derives disease treatment recommendations intelligently and accurately.

cs.LG↗

arXiv2018

Attention-Guided Answer Distillation for Machine Reading Comprehension

Despite that current reading comprehension systems have achieved significant advancements, their promising performances are often obtained at the cost of making an ensemble of numerous models. Besides, existing approaches are also vulnerable to adversarial attacks. This paper tackles these problems by leveraging knowledge distillation, which aims to transfer knowledge from an ensemble model to a single model. We first demonstrate that vanilla knowledge distillation applied to answer span prediction is effective for reading comprehension systems. We then propose two novel approaches that not only penalize the prediction on confusing answers but also guide the training with alignment information distilled from the ensemble. Experiments show that our best student model has only a slight drop of 0.4% F1 on the SQuAD test set compared to the ensemble teacher, while running 12x faster during inference. It even outperforms the teacher on adversarial SQuAD datasets and NarrativeQA benchmark.

cs.CL↗

arXiv2018

Joint Trajectory and Resource Allocation Design for UAV Communication Systems

In this paper, we investigate resource allocation design for unmanned aerial vehicle (UAV)-enabled communication systems, where a UAV is dispatched to provide communications to multiple user nodes. Our objective is to maximize the communication system throughput by jointly optimizing the subcarrier allocation policy and the trajectory of the UAV, while taking into account the minimum required data rate for each user node, no-fly zones (NFZs), the maximum UAV cruising speed, and initial/final UAV locations. The design is formulated as a mixed integer non-convex optimization problem which is generally intractable. Subsequently, a computationally-efficient iterative algorithm is proposed to obtain a locally optimal solution. Simulation results illustrate that the performance of the proposed iterative algorithm approaches closely to that of the system without NFZ. In addition, the proposed algorithm can achieve a significant throughput gain compared to various benchmark schemes.

cs.IT↗

arXiv2018

Deep Virtual Stereo Odometry: Leveraging Deep Depth Prediction for Monocular Direct Sparse Odometry

Monocular visual odometry approaches that purely rely on geometric cues are prone to scale drift and require sufficient motion parallax in successive frames for motion estimation and 3D reconstruction. In this paper, we propose to leverage deep monocular depth prediction to overcome limitations of geometry-based monocular visual odometry. To this end, we incorporate deep depth predictions into Direct Sparse Odometry (DSO) as direct virtual stereo measurements. For depth prediction, we design a novel deep network that refines predicted depth from a single image in a two-stage process. We train our network in a semi-supervised way on photoconsistency in stereo images and on consistency with accurate sparse depth reconstructions from Stereo DSO. Our deep predictions excel state-of-the-art approaches for monocular depth on the KITTI benchmark. Moreover, our Deep Virtual Stereo Odometry clearly exceeds previous monocular and deep learning based methods in accuracy. It even achieves comparable performance to the state-of-the-art stereo methods, while only relying on a single camera.

cs.CV↗