arXiv subjects

Xun Huang

Publications and source records attributed to Xun Huang.

51 records · Page 3Linked to original sources

arXiv2019

Finite Temperature Behaviors of q-deformed Fermi Gases

During the last three decades, non-standard statistics for indistinguishable quantum particles has attracted broad attentions and research interests from many institutions. Among these new types of statistics, the q-deformed Bose and Fermi statistics, originated from the study of quantum algebra, are being applied in more and more physical systems. In this paper, we construct a q-deformed generalization of the BCS-Leggett theory for ultracold Fermi gases based on our previously constructed q-deformed BCS theory. Some interesting features of this q-deformed interacting quantum gas are obtained by numerical analysis. For example, in the ordinary Bose-Einstein Condensation regime, the gas presents a fermionic feature instead of bosonic feature if the deformation parameter is tuned suitably, which might be referred to as the q-induced ``Bose-Fermi'' crossover. Conversely, a weak sign of the ``Fermi-Bose'' crossover is also found in the ordinary weak fermionic regime.

cond-mat.stat-mech↗

arXiv2019

Few-Shot Unsupervised Image-to-Image Translation

Unsupervised image-to-image translation methods learn to map images in a given class to an analogous image in a different class, drawing on unstructured (non-registered) datasets of images. While remarkably successful, current methods require access to many images in both source and destination classes at training time. We argue this greatly limits their use. Drawing inspiration from the human capability of picking up the essence of a novel object from a small number of examples and generalizing from there, we seek a few-shot, unsupervised image-to-image translation algorithm that works on previously unseen target classes that are specified, at test time, only by a few example images. Our model achieves this few-shot generation capability by coupling an adversarial training scheme with a novel network design. Through extensive experimental validation and comparisons to several baseline methods on benchmark datasets, we verify the effectiveness of the proposed framework. Our implementation and datasets are available at https://github.com/NVlabs/FUNIT .

cs.CV↗

arXiv2019

PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows

As 3D point clouds become the representation of choice for multiple vision and graphics applications, the ability to synthesize or reconstruct high-resolution, high-fidelity point clouds becomes crucial. Despite the recent success of deep learning models in discriminative tasks of point clouds, generating point clouds remains challenging. This paper proposes a principled probabilistic framework to generate 3D point clouds by modeling them as a distribution of distributions. Specifically, we learn a two-level hierarchy of distributions where the first level is the distribution of shapes and the second level is the distribution of points given a shape. This formulation allows us to both sample shapes and sample an arbitrary number of points from a shape. Our generative model, named PointFlow, learns each level of the distribution with a continuous normalizing flow. The invertibility of normalizing flows enables the computation of the likelihood during training and allows us to train our model in the variational inference framework. Empirically, we demonstrate that PointFlow achieves state-of-the-art performance in point cloud generation. We additionally show that our model can faithfully reconstruct point clouds and learn useful representations in an unsupervised manner. The code will be available at https://github.com/stevenygd/PointFlow.

cs.CV↗

arXiv2018

Multimodal Unsupervised Image-to-Image Translation

Unsupervised image-to-image translation is an important and challenging problem in computer vision. Given an image in the source domain, the goal is to learn the conditional distribution of corresponding images in the target domain, without seeing any pairs of corresponding images. While this conditional distribution is inherently multimodal, existing approaches make an overly simplified assumption, modeling it as a deterministic one-to-one mapping. As a result, they fail to generate diverse outputs from a given source domain image. To address this limitation, we propose a Multimodal Unsupervised Image-to-image Translation (MUNIT) framework. We assume that the image representation can be decomposed into a content code that is domain-invariant, and a style code that captures domain-specific properties. To translate an image to another domain, we recombine its content code with a random style code sampled from the style space of the target domain. We analyze the proposed framework and establish several theoretical results. Extensive experiments with comparisons to the state-of-the-art approaches further demonstrates the advantage of the proposed framework. Moreover, our framework allows users to control the style of translation outputs by providing an example style image. Code and pretrained models are available at https://github.com/nvlabs/MUNIT

cs.CV↗

arXiv2018

Learning to Evaluate Image Captioning

Evaluation metrics for image captioning face two challenges. Firstly, commonly used metrics such as CIDEr, METEOR, ROUGE and BLEU often do not correlate well with human judgments. Secondly, each metric has well known blind spots to pathological caption constructions, and rule-based metrics lack provisions to repair such blind spots once identified. For example, the newly proposed SPICE correlates well with human judgments, but fails to capture the syntactic structure of a sentence. To address these two challenges, we propose a novel learning based discriminative evaluation metric that is directly trained to distinguish between human and machine-generated captions. In addition, we further propose a data augmentation scheme to explicitly incorporate pathological transformations as negative examples during training. The proposed metric is evaluated with three kinds of robustness tests and its correlation with human judgments. Extensive experiments show that the proposed data augmentation scheme not only makes our metric more robust toward several pathological transformations, but also improves its correlation with human judgments. Our metric outperforms other metrics on both caption level human correlation in Flickr 8k and system level human correlation in COCO. The proposed approach could be served as a learning based evaluation metric that is complementary to existing rule-based metrics.

cs.CV↗

arXiv2018

$q$-deformed Fermion in Many-Particle Systems and Its Application to BCS Theory

In recent decades, there have been increasing interests in quantum statistics beyond the standard Fermi-Dirac and Bose-Einstein statistics, such as the fractional statistics, quon statistics, anyon statistics and quantum groups, since they can provide some new insights into the cosmology, nuclear physics and condensed matter. In this paper, we study the many-particle system formed by the $q$-deformed fermions ($q$-fermion), which is realized by deforming the quantum algebra of the anticommutation relations. We investigate from a standard perspective of the finite temperature field theory and try to construct the finite temperature Green's function formalism for the free many-$q$-fermion system, then generalize it to the well known interacting fermionic system, the superconductor, and finally obtain a consistent $q$-deformed BCS ($q$BCS) theory. At low temperature, this theory predicts a Sarma-like ordered phase, and we call it the $q$-deformed Sarma phase. It also presents a symmetric phase diagram in the parameter space and new thermodynamic relations.

cond-mat.stat-mech↗

arXiv2017

Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization

Gatys et al. recently introduced a neural algorithm that renders a content image in the style of another image, achieving so-called style transfer. However, their framework requires a slow iterative optimization process, which limits its practical application. Fast approximations with feed-forward neural networks have been proposed to speed up neural style transfer. Unfortunately, the speed improvement comes at a cost: the network is usually tied to a fixed set of styles and cannot adapt to arbitrary new styles. In this paper, we present a simple yet effective approach that for the first time enables arbitrary style transfer in real-time. At the heart of our method is a novel adaptive instance normalization (AdaIN) layer that aligns the mean and variance of the content features with those of the style features. Our method achieves speed comparable to the fastest existing approach, without the restriction to a pre-defined set of styles. In addition, our approach allows flexible user controls such as content-style trade-off, style interpolation, color & spatial controls, all using a single feed-forward neural network.

cs.CV↗

arXiv2017

Stacked Generative Adversarial Networks

In this paper, we propose a novel generative model named Stacked Generative Adversarial Networks (SGAN), which is trained to invert the hierarchical representations of a bottom-up discriminative network. Our model consists of a top-down stack of GANs, each learned to generate lower-level representations conditioned on higher-level representations. A representation discriminator is introduced at each feature hierarchy to encourage the representation manifold of the generator to align with that of the bottom-up discriminative network, leveraging the powerful discriminative representations to guide the generative model. In addition, we introduce a conditional loss that encourages the use of conditional information from the layer above, and a novel entropy loss that maximizes a variational lower bound on the conditional entropy of generator outputs. We first train each stack independently, and then train the whole model end-to-end. Unlike the original GAN that uses a single noise vector to represent all the variations, our SGAN decomposes variations into multiple levels and gradually resolves uncertainties in the top-down generative process. Based on visual inspection, Inception scores and visual Turing test, we demonstrate that SGAN is able to generate images of much higher quality than GANs without stacking.

cs.CV↗

arXiv2016

Top-Down Learning for Structured Labeling with Convolutional Pseudoprior

Current practice in convolutional neural networks (CNN) remains largely bottom-up and the role of top-down process in CNN for pattern analysis and visual inference is not very clear. In this paper, we propose a new method for structured labeling by developing convolutional pseudo-prior (ConvPP) on the ground-truth labels. Our method has several interesting properties: (1) compared with classical machine learning algorithms like CRFs and Structural SVM, ConvPP automatically learns rich convolutional kernels to capture both short- and long- range contexts; (2) compared with cascade classifiers like Auto-Context, ConvPP avoids the iterative steps of learning a series of discriminative classifiers and automatically learns contextual configurations; (3) compared with recent efforts combing CNN models with CRFs and RNNs, ConvPP learns convolution in the labeling space with much improved modeling capability and less manual specification; (4) compared with Bayesian models like MRFs, ConvPP capitalizes on the rich representation power of convolution by automatically learning priors built on convolutional filters. We accomplish our task using pseudo-likelihood approximation to the prior under a novel fixed-point network structure that facilitates an end-to-end learning process. We show state-of-the-art results on sequential labeling and image labeling benchmarks.

cs.CV↗

arXiv2014

Learning of Proto-object Representations via Fixations on Low Resolution

While previous researches in eye fixation prediction typically rely on integrating low-level features (e.g. color, edge) to form a saliency map, recently it has been found that the structural organization of these features into a proto-object representation can play a more significant role. In this work, we present a computational framework based on deep network to demonstrate that proto-object representations can be learned from low-resolution image patches from fixation regions. We advocate the use of low-resolution inputs in this work due to the following reasons: (1) Proto-objects are computed in parallel over an entire visual field (2) People can perceive or recognize objects well even it is in low resolution. (3) Fixations from lower resolution images can predict fixations on higher resolution images. In the proposed computational model, we extract multi-scale image patches on fixation regions from eye fixation datasets, resize them to low resolution and feed them into a hierarchical. With layer-wise unsupervised feature learning, we find that many proto-objects like features responsive to different shapes of object blobs are learned out. Visualizations also show that these features are selective to potential objects in the scene and the responses of these features work well in predicting eye fixations on the images when combined with learned weights.

cs.CV↗

arXiv2013

Acoustic Invisibility in Turbulent Fluids

Acoustic invisibility of a cloaking system in turbulent uids has been poorly understood. Here we show that evident scattering would appear in turbulent wakes owing to the submergence of a classical cloaking device. The inherent mechanism is explained using our theoretical model, which eventually inspires us to develop an optimized cloaking approach. Both the near- and far-field scatted fields are examined using high order computational acoustic methods. The remarkably low scattering demonstrates the effectiveness of the proposed acoustic cloaking approach for turbulent uid cases.

physics.flu-dyn↗

arXiv2013

Compressive sensing based beamforming for noisy measurements

Compressive sensing is the newly emerging method in information technology that could impact array beamforming and the associated engineering applications. However, practical measurements are inevitably polluted by noise from external interference and internal acquisition process. Then, compressive sensing based beamforming was studied in this work for those noisy measurements with a signal-to-noise ratio. In this article, we firstly introduced the fundamentals of compressive sensing theory. After that, we implemented two algorithms (CSB-I and CSB-II). Both algorithms are proposed for those presumably spatially sparse and incoherent signals. The two algorithms were examined using a simple simulation case and a practical aeroacoustic test case. The simulation case clearly shows that the CSB-I algorithm is quite sensitive to the sensing noise. The CSB-II algorithm, on the other hand, is more robust to noisy measurements. The results by CSB-II at $\mathrm{SNR}=-10\,$dB are still reasonable with good resolution and sidelobe rejection. Therefore, compressive sensing beamforming can be considered as a promising array signal beamforming method for those measurements with inevitably noisy interference.

cs.IT↗

arXiv2012

Circular Sound Wave Scattering Derivation for Acoustic Cloak Detection

In this Letter we develop analytical formulations to describe sound scattering in lossless medium due to 2D circular wave incident on an acoustic cloak. A perfect acoustic cloak is reflectionless and can completely hide the cloaked object from any sound waves. However, the realization of a perfect acoustic cloak is difficult. Compared to plane wave, our analytic calculations show that circular wave from an annular line source generates distinct scattering patterns from an imperfect cloak design. Large modification in reflection directivities can be observed if the focal point of the incident wavefront is slightly customized. Hence, our work might find applications in acoustic cloak detection, which should have significant impact on cloak design and defense.

physics.class-ph↗

arXiv2012

Lateral Stability Analysis of Hypersonic Vehicle under Pressure Fluctuation by Solving Mathieu Differential Equation

Two recent test failures of Hypersonic Technology Vehicle 2 impose a strike to the increasingly growing enthusiasm, not only on the United States side. It is important to find out the exact failure reason, otherwise a solution is impossible. In this Note, we propose a potential failure reason from the perspective of lateral stability analysis. We argue that the time variant pressure fluctuations, which are normally omitted in classical aircraft dynamics analysis, could not be neglected in dynamic analysis of hypersonic vehicles. To demonstrate the idea, a hypersonic model is imagined in this work and its aerodynamic parameters are estimated using fundamental fluid principles. Pressure fluctuations are thereafter estimated by an empirical formula. A lateral dynamic equation is set up, taking those time variant fluctuations into account. The resulted equation is a Mathieu differential equation. Numerical solutions of this equation show that the inclusion of fluctuation terms generates more complicated dynamics and should be considered in flight controller design.

physics.flu-dyn↗

arXiv2012

Airfoil Roll Control by Bang-Bang Optimal Control Method with Plasma Actuators

The bang-bang optimal control method was proposed for glow discharge plasma actuators, taking account of practical issues, such as limited actuation states with instantaneously varied aerodynamic control performance. Hence, the main contribution of this Note is to integrate flight control with active flow control in particular for plasma actuators. Flow control effects were examined in wind tunnel experiments, which show that the plasma authority for flow control is limited. Flow control effects are only obvious at pitch angles near stall. However, flight control simulations suggest that even those small plasma-induced roll moments can satisfactorily fulfill the maneuver tasks and meet flight quality specifications. In addition, the disturbance from volatile plasma-induced roll moments can be rejected. Hence, the proposed bang-bang control method is a promising candidate of control design methodology for plasma actuators.

physics.plasm-ph↗