arXiv subjects

Hao Zhang

Publications and source records attributed to Hao Zhang.

At least 1,153 records · Page 64Linked to original sources

arXiv2019

Learning Implicit Fields for Generative Shape Modeling

We advocate the use of implicit fields for learning generative models of shapes and introduce an implicit field decoder, called IM-NET, for shape generation, aimed at improving the visual quality of the generated shapes. An implicit field assigns a value to each point in 3D space, so that a shape can be extracted as an iso-surface. IM-NET is trained to perform this assignment by means of a binary classifier. Specifically, it takes a point coordinate, along with a feature vector encoding a shape, and outputs a value which indicates whether the point is outside the shape or not. By replacing conventional decoders by our implicit decoder for representation learning (via IM-AE) and shape generation (via IM-GAN), we demonstrate superior results for tasks such as generative shape modeling, interpolation, and single-view 3D reconstruction, particularly in terms of visual quality. Code and supplementary material are available at https://github.com/czq142857/implicit-decoder.

cs.GR↗

arXiv2019

SDM-NET: Deep Generative Network for Structured Deformable Mesh

We introduce SDM-NET, a deep generative neural network which produces structured deformable meshes. Specifically, the network is trained to generate a spatial arrangement of closed, deformable mesh parts, which respect the global part structure of a shape collection, e.g., chairs, airplanes, etc. Our key observation is that while the overall structure of a 3D shape can be complex, the shape can usually be decomposed into a set of parts, each homeomorphic to a box, and the finer-scale geometry of the part can be recovered by deforming the box. The architecture of SDM-NET is that of a two-level variational autoencoder (VAE). At the part level, a PartVAE learns a deformable model of part geometries. At the structural level, we train a Structured Parts VAE (SP-VAE), which jointly learns the part structure of a shape collection and the part geometries, ensuring a coherence between global shape structure and surface details. Through extensive experiments and comparisons with the state-of-the-art deep generative models of shapes, we demonstrate the superiority of SDM-NET in generating meshes with visual quality, flexible topology, and meaningful structures, which benefit shape interpolation and other subsequently modeling tasks.

cs.GR↗

arXiv2019

LOGAN: Unpaired Shape Transform in Latent Overcomplete Space

We introduce LOGAN, a deep neural network aimed at learning general-purpose shape transforms from unpaired domains. The network is trained on two sets of shapes, e.g., tables and chairs, while there is neither a pairing between shapes from the domains as supervision nor any point-wise correspondence between any shapes. Once trained, LOGAN takes a shape from one domain and transforms it into the other. Our network consists of an autoencoder to encode shapes from the two input domains into a common latent space, where the latent codes concatenate multi-scale shape features, resulting in an overcomplete representation. The translator is based on a generative adversarial network (GAN), operating in the latent space, where an adversarial loss enforces cross-domain translation while a feature preservation loss ensures that the right shape features are preserved for a natural shape transform. We conduct ablation studies to validate each of our key network designs and demonstrate superior capabilities in unpaired shape transforms on a variety of examples over baselines and state-of-the-art approaches. We show that LOGAN is able to learn what shape features to preserve during shape translation, either local or non-local, whether content or style, depending solely on the input domains for training.

cs.CV↗

arXiv2019

CompoNet: Learning to Generate the Unseen by Part Synthesis and Composition

Data-driven generative modeling has made remarkable progress by leveraging the power of deep neural networks. A reoccurring challenge is how to enable a model to generate a rich variety of samples from the entire target distribution, rather than only from a distribution confined to the training data. In other words, we would like the generative model to go beyond the observed samples and learn to generate ``unseen'', yet still plausible, data. In our work, we present CompoNet, a generative neural network for 2D or 3D shapes that is based on a part-based prior, where the key idea is for the network to synthesize shapes by varying both the shape parts and their compositions. Treating a shape not as an unstructured whole, but as a (re-)composable set of deformable parts, adds a combinatorial dimension to the generative process to enrich the diversity of the output, encouraging the generator to venture more into the ``unseen''. We show that our part-based model generates richer variety of plausible shapes compared with baseline generative models. To this end, we introduce two quantitative metrics to evaluate the diversity of a generative model and assess how well the generated data covers both the training data and unseen data from the same target distribution. Code is available at https://github.com/nschor/CompoNet.

cs.GR↗

arXiv2019

Phononic entanglement concentration via optomechanical interactions

Low dissipation, tunable coupling to other quantum systems, and unique features of phonons in the aspects of propagation, detection and others suggest the applications of quantized mechanical resonators in phonon-based quantum information processing (QIP) in a way different from their photonic counterpart. In this paper, we propose the first protocol of entanglement concentration for nonlocal phonons from quantized mechanical vibration. We combine the optomechanical cross-Kerr interaction with the Mach-Zehnder interferometer and, by means of twice optomechanical interactions and the photon analysis with respect to the output of the interferometer, achieve ideal entanglement concentration about less-entangled nonlocal phonon Bell and Greenberger-Horne-Zeilinger states. Our protocol is useful for preserving the entangled phonons for the use of high quality phonon-based QIP in future.

quant-ph↗

arXiv2019

An End-to-End Neural Network for Image Cropping by Learning Composition from Aesthetic Photos

As one of the fundamental techniques for image editing, image cropping discards unrelevant contents and remains the pleasing portions of the image to enhance the overall composition and achieve better visual/aesthetic perception. In this paper, we primarily focus on improving the accuracy of automatic image cropping, and on further exploring its potential in public datasets with high efficiency. From this respect, we propose a deep learning based framework to learn the objects composition from photos with high aesthetic qualities, where an anchor region is detected through a convolutional neural network (CNN) with the Gaussian kernel to maintain the interested objects' integrity. This initial detected anchor area is then fed into a light weighted regression network to obtain the final cropping result. Unlike the conventional methods that multiple candidates are proposed and evaluated iteratively, only a single anchor region is produced in our model, which is mapped to the final output directly. Thus, low computational resources are required for the proposed approach. The experimental results on the public datasets show that both cropping accuracy and efficiency achieve the state-ofthe-art performances.

cs.CV↗

arXiv2019

Generating an Overview Report over Many Documents

How to efficiently generate an accurate, well-structured overview report (ORPT) over thousands of related documents is challenging. A well-structured ORPT consists of sections of multiple levels (e.g., sections and subsections). None of the existing multi-document summarization (MDS) algorithms is directed toward this task. To overcome this obstacle, we present NDORGS (Numerous Documents' Overview Report Generation Scheme) that integrates text filtering, keyword scoring, single-document summarization (SDS), topic modeling, MDS, and title generation to generate a coherent, well-structured ORPT. We then devise a multi-criteria evaluation method using techniques of text mining and multi-attribute decision making on a combination of human judgments, running time, information coverage, and topic diversity. We evaluate ORPTs generated by NDORGS on two large corpora of documents, where one is classified and the other unclassified. We show that, using Saaty's pairwise comparison 9-point scale and under TOPSIS, the ORPTs generated on SDS's with the length of 20% of the original documents are the best overall on both datasets.

cs.CL↗

arXiv2019

BAE-NET: Branched Autoencoder for Shape Co-Segmentation

We treat shape co-segmentation as a representation learning problem and introduce BAE-NET, a branched autoencoder network, for the task. The unsupervised BAE-NET is trained with a collection of un-segmented shapes, using a shape reconstruction loss, without any ground-truth labels. Specifically, the network takes an input shape and encodes it using a convolutional neural network, whereas the decoder concatenates the resulting feature code with a point coordinate and outputs a value indicating whether the point is inside/outside the shape. Importantly, the decoder is branched: each branch learns a compact representation for one commonly recurring part of the shape collection, e.g., airplane wings. By complementing the shape reconstruction loss with a label loss, BAE-NET is easily tuned for one-shot learning. We show unsupervised, weakly supervised, and one-shot learning results by BAE-NET, demonstrating that using only a couple of exemplars, our network can generally outperform state-of-the-art supervised methods trained on hundreds of segmented shapes. Code is available at https://github.com/czq142857/BAE-NET.

cs.CV↗

arXiv2019

Improving Performance of End-to-End ASR on Numeric Sequences

Recognizing written domain numeric utterances (e.g. I need $1.25.) can be challenging for ASR systems, particularly when numeric sequences are not seen during training. This out-of-vocabulary (OOV) issue is addressed in conventional ASR systems by training part of the model on spoken domain utterances (e.g. I need one dollar and twenty five cents.), for which numeric sequences are composed of in-vocabulary numbers, and then using an FST verbalizer to denormalize the result. Unfortunately, conventional ASR models are not suitable for the low memory setting of on-device speech recognition. E2E models such as RNN-T are attractive for on-device ASR, as they fold the AM, PM and LM of a conventional model into one neural network. However, in the on-device setting the large memory footprint of an FST denormer makes spoken domain training more difficult. In this paper, we investigate techniques to improve E2E model performance on numeric data. We find that using a text-to-speech system to generate additional numeric training data, as well as using a small-footprint neural network to perform spoken-to-written domain denorming, yields improvement in several numeric classes. In the case of the longest numeric sequences, we see reduction of WER by up to a factor of 8.

eess.AS↗

arXiv2019

Variable optical chirality in an atomic assisted microresonator

A whispering gallery mode resonator with a cavity-made slot filled atomic vapor is demonstrated, in which the chiral symmetry is broken induced by asymmetric backscattering of counter-propagating optical waves in the WGM microcavity. The chirality of resonator modes is steered via tuning dispersion relation in the cavity-made slot. The system exhibiting a high displacement sensitivity 15.22 THz/nm could be applied to optomechanical sensing applications.

physics.optics↗

arXiv2019

Compressive spectral imaging based on hexagonal blue noise coded apertures

Coded aperture snapshot spectral imager (CASSI) is a computational imaging system that acquires a three dimensional (3D) spectral data cube by single or a few two dimensional (2D) measurements. Binary random coded apertures with square pixels are primarily implemented in CASSI systems to modulate the spectral images in spatial domain. The design and optimization of coded apertures was shown to improve the imaging performance of these systems significantly. This work proposes a different approach to code design. Instead of traditional squared tiled coded elements, hexagonal tiled elements are used. The dislocation between the binary hexagonal pixels on coded apertures and the square pixels on detector introduces equivalent grey-scale spatial modulation to increase the degrees of freedom in the sensing matrix, thus further improving the spectral imaging performance. Then, this paper presents an optimal structure under the criterion of satisfying the restricted isometry property (RIP) with high probability, coined blue noise (BN) coded apertures. In addition, this paper studies and verifies the proposed hexagonal blue noise coded aperture method on a general CASSI system, where the resolution of the coded aperture is equivalent to that of the detector. Based on the RIP criterion, this paper theoretically proves the superiority of the hexagonal blue noise coded aperture over the traditional random coded aperture with square lattice.

eess.IV↗

arXiv2019

Constrained low-tubal-rank tensor recovery for hyperspectral images mixed noise removal by bilateral random projections

In this paper, we propose a novel low-tubal-rank tensor recovery model, which directly constrains the tubal rank prior for effectively removing the mixed Gaussian and sparse noise in hyperspectral images. The constraints of tubal-rank and sparsity can govern the solution of the denoised tensor in the recovery procedure. To solve the constrained low-tubal-rank model, we develop an iterative algorithm based on bilateral random projections to efficiently solve the proposed model. The advantage of random projections is that the approximation of the low-tubal-rank tensor can be obtained quite accurately in an inexpensive manner. Experimental examples for hyperspectral image denoising are presented to demonstrate the effectiveness and efficiency of the proposed method.

cs.CV↗

arXiv2019

Spin-Orbit Protection of Induced Superconductivity in Majorana Nanowires

Spin-orbit interaction (SOI) plays a key role in creating Majorana zero modes in semiconductor nanowires proximity coupled to a superconductor. We track the evolution of the induced superconducting gap in InSb nanowires coupled to a NbTiN superconductor in a large range of magnetic field strengths and orientations. Based on realistic simulations of our devices, we reveal SOI with a strength of 0.15-0.35 eV$\require{mediawiki-texvc}Å$. Our approach identifies the direction of the spin-orbit field, which is strongly affected by the superconductor geometry and electrostatic gates.

cond-mat.mes-hall↗

arXiv2019

UKP-Athene: Multi-Sentence Textual Entailment for Claim Verification

The Fact Extraction and VERification (FEVER) shared task was launched to support the development of systems able to verify claims by extracting supporting or refuting facts from raw text. The shared task organizers provide a large-scale dataset for the consecutive steps involved in claim verification, in particular, document retrieval, fact extraction, and claim classification. In this paper, we present our claim verification pipeline approach, which, according to the preliminary results, scored third in the shared task, out of 23 competing systems. For the document retrieval, we implemented a new entity linking approach. In order to be able to rank candidate facts and classify a claim on the basis of several selected facts, we introduce two extensions to the Enhanced LSTM (ESIM).

cs.IR↗

arXiv2019

GRAINS: Generative Recursive Autoencoders for INdoor Scenes

We present a generative neural network which enables us to generate plausible 3D indoor scenes in large quantities and varieties, easily and highly efficiently. Our key observation is that indoor scene structures are inherently hierarchical. Hence, our network is not convolutional; it is a recursive neural network or RvNN. Using a dataset of annotated scene hierarchies, we train a variational recursive autoencoder, or RvNN-VAE, which performs scene object grouping during its encoding phase and scene generation during decoding. Specifically, a set of encoders are recursively applied to group 3D objects based on support, surround, and co-occurrence relations in a scene, encoding information about object spatial properties, semantics, and their relative positioning with respect to other objects in the hierarchy. By training a variational autoencoder (VAE), the resulting fixed-length codes roughly follow a Gaussian distribution. A novel 3D scene can be generated hierarchically by the decoder from a randomly sampled code from the learned distribution. We coin our method GRAINS, for Generative Recursive Autoencoders for INdoor Scenes. We demonstrate the capability of GRAINS to generate plausible and diverse 3D indoor scenes and compare with existing methods for 3D scene synthesis. We show applications of GRAINS including 3D scene modeling from 2D layouts, scene editing, and semantic scene segmentation via PointNet whose performance is boosted by the large quantity and variety of 3D scenes generated by our method.

cs.GR↗

arXiv2019

DenseAttentionSeg: Segment Hands from Interacted Objects Using Depth Input

We propose a real-time DNN-based technique to segment hand and object of interacting motions from depth inputs. Our model is called DenseAttentionSeg, which contains a dense attention mechanism to fuse information in different scales and improves the results quality with skip-connections. Besides, we introduce a contour loss in model training, which helps to generate accurate hand and object boundaries. Finally, we propose and release our InterSegHands dataset, a fine-scale hand segmentation dataset containing about 52k depth maps of hand-object interactions. Our experiments evaluate the effectiveness of our techniques and datasets, and indicate that our method outperforms the current state-of-the-art deep segmentation methods on interaction segmentation.

cs.CV↗

arXiv2019

Decays of Majorana or Andreev oscillations induced by steplike spin-orbit coupling

The Majorana zero mode in the semiconductor-superconductor nanowire is one of the promising candidates for topological quantum computing. Recently, in islands of nanowires, subgap-state energies have been experimentally observed to oscillate as a function of the magnetic field, showing a signature of overlapped Majorana bound states. However, the oscillation amplitude either dies away after an overshoot or decays, sharply opposite to the theoretically predicted enhanced oscillations for Majorana bound states. We reveal that a steplike distribution of spin-orbit coupling in realistic devices can induce the decaying Majorana oscillations, resulting from the coupling-induced energy repulsion between the quasiparticle spectra on the two sides of the step. This steplike spin-orbit coupling can also lead to decaying oscillations in the spectrum of the Andreev bound states. For Coulomb-blockade peaks mediated by the Majorana bound states, the peak spacings have been predicted to correlate with peak heights by a $π/2$ phase shift, which was ambiguous in recent experiments and may be explained by the steplike spin-orbit coupling. Our work will inspire more works to reexamine effects of the nonuniform spin-orbit coupling, which is generally present in experimental devices.

cond-mat.mes-hall↗

arXiv2019

Arbitrary function resonance tuner of the optical microcavity with sub-MHz resolution

The resonance frequency of an optical whispering gallery mode (WGM) microcavity is extremely important in its various applications. Many efforts have been made to fine tune this parameter. Here, we report the design and implementation of a function resonance tuner of an optical microcavity with resolution about 650 kHz (7 pm @ 1450 nm band), 20% of the optical WGM linewidth. A piezoelectric nano-positioner is used to mechanically compress the microsphere in its axial direction. The ultrafine frequency tuning is achieved benefitting from the much less changes in the axial direction than equatorial semiaxes of the microsphere and the sub-nanometer resolution of the nano-positioner. The tuning of the resonance can be made to an arbitrary function, dynamically, with near perfect accuracy. We have demonstrated the periodically tuning of resonance in the sine and sigmoid function respectively, both with over 99% fitting accuracy. This work expands the application of microresonators greatly, especially microspheres with ultrahigh quality factor, in multi-mode coupling system or time-floquet system.

physics.ins-det↗