arXiv subjects

Xiang Gao

Publications and source records attributed to Xiang Gao.

At least 271 records · Page 15Linked to original sources

arXiv2020

DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation

We present a large, tunable neural conversational response generation model, DialoGPT (dialogue generative pre-trained transformer). Trained on 147M conversation-like exchanges extracted from Reddit comment chains over a period spanning from 2005 through 2017, DialoGPT extends the Hugging Face PyTorch transformer to attain a performance close to human both in terms of automatic and human evaluation in single-turn dialogue settings. We show that conversational systems that leverage DialoGPT generate more relevant, contentful and context-consistent responses than strong baseline systems. The pre-trained model and training pipeline are publicly released to facilitate research into neural response generation and the development of more intelligent open-domain dialogue systems.

cs.CL↗

arXiv2020

3D Dynamic Point Cloud Denoising via Spatial-Temporal Graph Learning

The prevalence of accessible depth sensing and 3D laser scanning techniques has enabled the convenient acquisition of 3D dynamic point clouds, which provide efficient representation of arbitrarily-shaped objects in motion. Nevertheless, dynamic point clouds are often perturbed by noise due to hardware, software or other causes. While a plethora of methods have been proposed for static point cloud denoising, few efforts are made for the denoising of dynamic point clouds with varying number of irregularly-sampled points in each frame. In this paper, we represent dynamic point clouds naturally on graphs and address the denoising problem by inferring the underlying graph via spatio-temporal graph learning, exploiting both the intra-frame similarity and inter-frame consistency. Firstly, assuming the availability of a relevant feature vector per node, we pose spatial-temporal graph learning as optimizing a Mahalanobis distance metric $\mathbf{M}$, which is formulated as the minimization of graph Laplacian regularizer. Secondly, to ease the optimization of the symmetric and positive definite metric matrix $\mathbf{M}$, we decompose it into $\mathbf{M}=\mathbf{R}^{\top}\mathbf{R}$ and solve $\mathbf{R}$ instead via proximal gradient. Finally, based on the spatial-temporal graph learning, we formulate dynamic point cloud denoising as the joint optimization of the desired point cloud and underlying spatio-temporal graph, which leverages both intra-frame affinities and inter-frame consistency and is solved via alternating minimization. Experimental results show that the proposed method significantly outperforms independent denoising of each frame from state-of-the-art static point cloud denoising approaches.

cs.CV↗

arXiv2020

Pion valence quark PDF from lattice QCD

We present lattice results on the valence-quark structure of the pion using a coordinate space method within the framework of Large Momentum Effective Theory (LaMET). In this method one relies on the matrix elements of a Euclidean correlator in boosted hadronic states, which have an operator product expansion at short distance that allows us to extract the moments of PDFs. We renormalize the Euclidean correlator by forming the reduced Ioffe-time distribution (rITD), and reconstruct the second and fourth moments of the pion PDF by taking into account of QCD evolution effects.

hep-lat↗

arXiv2020

GraphTER: Unsupervised Learning of Graph Transformation Equivariant Representations via Auto-Encoding Node-wise Transformations

Recent advances in Graph Convolutional Neural Networks (GCNNs) have shown their efficiency for non-Euclidean data on graphs, which often require a large amount of labeled data with high cost. It it thus critical to learn graph feature representations in an unsupervised manner in practice. To this end, we propose a novel unsupervised learning of Graph Transformation Equivariant Representations (GraphTER), aiming to capture intrinsic patterns of graph structure under both global and local transformations. Specifically, we allow to sample different groups of nodes from a graph and then transform them node-wise isotropically or anisotropically. Then, we self-train a representation encoder to capture the graph structures by reconstructing these node-wise transformations from the feature representations of the original and transformed graphs. In experiments, we apply the learned GraphTER to graphs of 3D point cloud data, and results on point cloud segmentation/classification show that GraphTER significantly outperforms state-of-the-art unsupervised approaches and pushes greatly closer towards the upper bound set by the fully supervised counterparts. The code is available at: https://github.com/gyshgx868/graph-ter.

cs.CV↗

arXiv2020

Super Resolution Convolutional Neural Network for Feature Extraction in Spectroscopic Data

Two dimensional (2D) peak finding is a common practice in data analysis for physics experiments, which is typically achieved by computing the local derivatives. However, this method is inherently unstable when the local landscape is complicated, or the signal-to-noise ratio of the data is low. In this work, we propose a new method in which the peak tracking task is formalized as an inverse problem, thus can be solved with a convolutional neural network (CNN). In addition, we show that the underlying physics principle of the experiments can be used to generate the training data. By generalizing the trained neural network on real experimental data, we show that the CNN method can achieve comparable or better results than traditional derivative based methods. This approach can be further generalized in different physics experiments when the physical process is known.

physics.data-an↗

arXiv2020

Feature Graph Learning for 3D Point Cloud Denoising

Identifying an appropriate underlying graph kernel that reflects pairwise similarities is critical in many recent graph spectral signal restoration schemes, including image denoising, dequantization, and contrast enhancement. Existing graph learning algorithms compute the most likely entries of a properly defined graph Laplacian matrix $\mathbf{L}$, but require a large number of signal observations $\mathbf{z}$'s for a stable estimate. In this work, we assume instead the availability of a relevant feature vector $\mathbf{f}_i$ per node $i$, from which we compute an optimal feature graph via optimization of a feature metric. Specifically, we alternately optimize the diagonal and off-diagonal entries of a Mahalanobis distance matrix $\mathbf{M}$ by minimizing the graph Laplacian regularizer (GLR) $\mathbf{z}^{\top} \mathbf{L} \mathbf{z}$, where edge weight is $w_{i,j} = \exp\{-(\mathbf{f}_i - \mathbf{f}_j)^{\top} \mathbf{M} (\mathbf{f}_i - \mathbf{f}_j) \}$, given a single observation $\mathbf{z}$. We optimize diagonal entries via proximal gradient (PG), where we constrain $\mathbf{M}$ to be positive definite (PD) via linear inequalities derived from the Gershgorin circle theorem. To optimize off-diagonal entries, we design a block descent algorithm that iteratively optimizes one row and column of $\mathbf{M}$. To keep $\mathbf{M}$ PD, we constrain the Schur complement of sub-matrix $\mathbf{M}_{2,2}$ of $\mathbf{M}$ to be PD when optimizing via PG. Our algorithm mitigates full eigen-decomposition of $\mathbf{M}$, thus ensuring fast computation speed even when feature vector $\mathbf{f}_i$ has high dimension. To validate its usefulness, we apply our feature graph learning algorithm to the problem of 3D point cloud denoising, resulting in state-of-the-art performance compared to competing schemes in extensive experiments.

cs.CV↗

arXiv2019

Dynamic and programmable cellular-scale granules enable tissue-like materials

Tissue-like materials are required in many robotic systems to improve human-machine interactions. However, the mechanical properties of living tissues are difficult to replicate. Synthetic materials are not usually capable of simultaneously displaying the behaviors of the cellular ensemble and the extracellular matrix. A particular challenge is identification of a cell-like synthetic component which is tightly integrated with its matrix and also responsive to external stimuli at the population level. Here, we demonstrate that cellular-scale hydrated starch granules, an underexplored component in materials science, can turn conventional hydrogels into tissue-like materials when composites are formed. Using several synchrotron-based X-ray techniques, we reveal the mechanically-induced motion and training dynamics of the starch granules in the hydrogel matrix. These dynamic behaviors enable multiple tissue-like properties such as strain-stiffening, anisotropy, mechanical heterogeneity, programmability, mechanochemistry, impact absorption, and self-healability. The starch-hydrogel composites can be processed as robotic skins that maintain these tissue-like characteristics.

cond-mat.soft↗

arXiv2019

Strong Modification of Radiative Transition Rates due to a Breit-Interaction-Induced Avoided Crossing

We present the observations of x-rays emitted from the $1s2s^{2}2p_{1/2}2p_{3/2}$ inner shell excited state of B-like W and Bi ions. The relative transition rates are obtained for two dominant radiative transitions to $1s^{2}2s^{2}2p_{1/2}$ and $1s^{2}2s^{2}2p_{3/2}$. The experimental results and the comparison with rigorous relativistic calculations show that the rates of the strong electric dipole allowed $1s^22s^22p$ -- $1s2s^22p^2$ transitions are strongly modified due to a drastic change in the wavefunction caused by the Breit interaction.

physics.atom-ph↗

arXiv2019

Complete Scene Reconstruction by Merging Images and Laser Scans

Image based modeling and laser scanning are two commonly used approaches in large-scale architectural scene reconstruction nowadays. In order to generate a complete scene reconstruction, an effective way is to completely cover the scene using ground and aerial images, supplemented by laser scanning on certain regions with low texture and complicated structure. Thus, the key issue is to accurately calibrate cameras and register laser scans in a unified framework. To this end, we proposed a three-step pipeline for complete scene reconstruction by merging images and laser scans. First, images are captured around the architecture in a multi-view and multi-scale way and are feed into a structure-from-motion (SfM) pipeline to generate SfM points. Then, based on the SfM result, the laser scanning locations are automatically planned by considering textural richness, structural complexity of the scene and spatial layout of the laser scans. Finally, the images and laser scans are accurately merged in a coarse-to-fine manner. Experimental evaluations on two ancient Chinese architecture datasets demonstrate the effectiveness of our proposed complete scene reconstruction pipeline.

cs.CV↗

arXiv2019

Exploring Structure-Adaptive Graph Learning for Robust Semi-Supervised Classification

Graph Convolutional Neural Networks (GCNNs) are generalizations of CNNs to graph-structured data, in which convolution is guided by the graph topology. In many cases where graphs are unavailable, existing methods manually construct graphs or learn task-driven adaptive graphs. In this paper, we propose Graph Learning Neural Networks (GLNNs), which exploit the optimization of graphs (the adjacency matrix in particular) from both data and tasks. Leveraging on spectral graph theory, we propose the objective of graph learning from a sparsity constraint, properties of a valid adjacency matrix as well as a graph Laplacian regularizer via maximum a posteriori estimation. The optimization objective is then integrated into the loss function of the GCNN, which adapts the graph topology to not only labels of a specific task but also the input data. Experimental results show that our proposed GLNN outperforms state-of-the-art approaches over widely adopted social network datasets and citation network datasets for semi-supervised classification.

cs.LG↗

arXiv2019

Joint Learning of Graph Representation and Node Features in Graph Convolutional Neural Networks

Graph Convolutional Neural Networks (GCNNs) extend classical CNNs to graph data domain, such as brain networks, social networks and 3D point clouds. It is critical to identify an appropriate graph for the subsequent graph convolution. Existing methods manually construct or learn one fixed graph for all the layers of a GCNN. In order to adapt to the underlying structure of node features in different layers, we propose dynamic learning of graphs and node features jointly in GCNNs. In particular, we cast the graph optimization problem as distance metric learning to capture pairwise similarities of features in each layer. We deploy the Mahalanobis distance metric and further decompose the metric matrix into a low-dimensional matrix, which converts graph learning to the optimization of a low-dimensional matrix for efficient implementation. Extensive experiments on point clouds and citation network datasets demonstrate the superiority of the proposed method in terms of both accuracies and robustness.

cs.LG↗

arXiv2019

Structuring Latent Spaces for Stylized Response Generation

Generating responses in a targeted style is a useful yet challenging task, especially in the absence of parallel data. With limited data, existing methods tend to generate responses that are either less stylized or less context-relevant. We propose StyleFusion, which bridges conversation modeling and non-parallel style transfer by sharing a structured latent space. This structure allows the system to generate stylized relevant responses by sampling in the neighborhood of the conversation model prediction, and continuously control the style level. We demonstrate this method using dialogues from Reddit data and two sets of sentences with distinct styles (arXiv and Sherlock Holmes novels). Automatic and human evaluation show that, without sacrificing appropriateness, the system generates responses of the targeted style and outperforms competitive baselines.

cs.CL↗

arXiv2019

Competing Phases in Epitaxial Vanadium Dioxide at Nanoscale

Phase competition in correlated oxides offers tantalizing opportunities as many intriguing physical phenomena occur near the phase transitions. Owing to a sharp metal-insulator transition (MIT) near room temperature, correlated vanadium dioxide (VO2) exhibits a strong competition between insulating and metallic phases that is important for practical applications. However, the phase boundary undergoes strong modification when strain is involved, yielding complex phase transitions. Here, we report the emergence of the nanoscale M2 phase domains in VO2 epitaxial films under anisotropic strain relaxation. The phase states of the films are imaged by multi-length-scale probes, detecting the structural and electrical properties in individual local domains. Competing evolution of the M1 and M2 phases indicates a critical role of lattice-strain on both the stability of the M2 Mott phase and the energetics of the MIT in VO2 films. This study demonstrates how strain engineering can be utilized to design phase states, which allow deliberate control of MIT behavior at the nanoscale in epitaxial VO2 films.

cond-mat.str-el↗

arXiv2019

Revealing Energy Dependence of Quantum Defects via Two Heteronuclear Atoms in an Optical Tweezer

As a physically motivated and computationally simple model for cold atomic and molecular collisions, the multichannel quantum defect theory (MQDT) with frame transformation (FT) formalism provides an analytical treatment of scattering resonances in an arbitrary partial wave between alkali-metal atoms, leading to the experimental observation of $p-$ and $d-$wave resonances. However, the inconsistency of quantum defects for describing scattering resonances shows up when compared with experiments. Here, with two heteronuclear atoms in the ground state of an optical tweezer, the energy dependence of quantum defects is obviously revealed by comparing the measured s-wave scattering length with the prediction of MQDT-FT. By dividing the quantum defects into energy sensitive and insensitive categories, the inconsistency is ultimately removed while retaining the analytic structure of MQDT-FT. This study represents a significant improvement in the analytical MQDT-FT and demonstrates that a clean two-particle system is valuable to the test of collisional physics.

quant-ph↗

arXiv2019

Conversing by Reading: Contentful Neural Conversation with On-demand Machine Reading

Although neural conversation models are effective in learning how to produce fluent responses, their primary challenge lies in knowing what to say to make the conversation contentful and non-vacuous. We present a new end-to-end approach to contentful neural conversation that jointly models response generation and on-demand machine reading. The key idea is to provide the conversation model with relevant long-form text on the fly as a source of external knowledge. The model performs QA-style reading comprehension on this text in response to each conversational turn, thereby allowing for more focused integration of external knowledge than has been possible in prior approaches. To support further research on knowledge-grounded conversation, we introduce a new large-scale conversation dataset grounded in external web pages (2.8M turns, 7.4M sentences of grounding). Both human evaluation and automated metrics show that our approach results in more contentful responses compared to a variety of previous methods, improving both the informativeness and diversity of generated output.

cs.CL↗

arXiv2019

Visual Localization Using Sparse Semantic 3D Map

Accurate and robust visual localization under a wide range of viewing condition variations including season and illumination changes, as well as weather and day-night variations, is the key component for many computer vision and robotics applications. Under these conditions, most traditional methods would fail to locate the camera. In this paper we present a visual localization algorithm that combines structure-based method and image-based method with semantic information. Given semantic information about the query and database images, the retrieved images are scored according to the semantic consistency of the 3D model and the query image. Then the semantic matching score is used as weight for RANSAC's sampling and the pose is solved by a standard PnP solver. Experiments on the challenging long-term visual localization benchmark dataset demonstrate that our method has significant improvement compared with the state-of-the-arts.

cs.CV↗

arXiv2019

A Multi-Task Learning Framework for Overcoming the Catastrophic Forgetting in Automatic Speech Recognition

Recently, data-driven based Automatic Speech Recognition (ASR) systems have achieved state-of-the-art results. And transfer learning is often used when those existing systems are adapted to the target domain, e.g., fine-tuning, retraining. However, in the processes, the system parameters may well deviate too much from the previously learned parameters. Thus, it is difficult for the system training process to learn knowledge from target domains meanwhile not forgetting knowledge from the previous learning process, which is called as catastrophic forgetting (CF). In this paper, we attempt to solve the CF problem with the lifelong learning and propose a novel multi-task learning (MTL) training framework for ASR. It considers reserving original knowledge and learning new knowledge as two independent tasks, respectively. On the one hand, we constrain the new parameters not to deviate too far from the original parameters and punish the new system when forgetting original knowledge. On the other hand, we force the new system to solve new knowledge quickly. Then, a MTL mechanism is employed to get the balance between the two tasks. We applied our method to an End2End ASR task and obtained the best performance in both target and original datasets.

cs.SD↗

arXiv2019

Variation of the transition energies and oscillator strengths for the 3C and 3D lines of the Ne-like ions under plasma environment

We present the results of a detailed theoretical study which meets the spatial and temporal criteria of the Debye-Huckel (DH) approximation on the variation of the transition energies as well as the oscillator strengths for the ${2p^53d\ ^1P_1\rightarrow2p^6\ ^1S_0}$ (3C line) and the ${2p^53d\ ^3D_1\rightarrow2p^6\ ^1S_0}$ (3D line) transitions of the Ne-like ions subject to external plasma.\ Our study shows that the redshifts of the transition energy follow the general scaling behaviors similar to the ones for the simple H-like and He-like ions.\ Whereas the oscillator strength for the 3C line decreases, the oscillator strength for the spin-flipped 3D line increases as the strength of the outside plasma increases.\ As a result, their ratio is amplified subject to outside plasma environment.\ We further demonstrate that the plasma-induced variation between the relative strength of the 3C and 3D transitions is mainly due to the spin-dependent interactions which dictate the mixing of the $^1P_1$ component in the $^3D_1$ upper state of the 3D transition.\ In addition, we are able to find that the ratio between the relative oscillator strengths of the 3C and 3D lines in the presence of the plasma to their respective plasma-free values varies as a nearly universal function of $[(Z-9.2)DZ]^{-1.8}$, with $Z$ the nuclear charge and $D$ the Debye length.\ The results of this study should be of great help in the modeling and diagnostic of astrophysical plasmas as well as laboratory plasmas.

physics.atom-ph↗