arXiv subjects

Sheng Huang

Publications and source records attributed to Sheng Huang.

At least 55 records · Page 3Linked to original sources

arXiv2022

An Iteratively Optimized Patch Label Inference Network for Automatic Pavement Distress Detection

We present a novel deep learning framework named the Iteratively Optimized Patch Label Inference Network (IOPLIN) for automatically detecting various pavement distresses that are not solely limited to specific ones, such as cracks and potholes. IOPLIN can be iteratively trained with only the image label via the Expectation-Maximization Inspired Patch Label Distillation (EMIPLD) strategy, and accomplish this task well by inferring the labels of patches from the pavement images. IOPLIN enjoys many desirable properties over the state-of-the-art single branch CNN models such as GoogLeNet and EfficientNet. It is able to handle images in different resolutions, and sufficiently utilize image information particularly for the high-resolution ones, since IOPLIN extracts the visual features from unrevised image patches instead of the resized entire image. Moreover, it can roughly localize the pavement distress without using any prior localization information in the training phase. In order to better evaluate the effectiveness of our method in practice, we construct a large-scale Bituminous Pavement Disease Detection dataset named CQU-BPDD consisting of 60,059 high-resolution pavement images, which are acquired from different areas at different times. Extensive results on this dataset demonstrate the superiority of IOPLIN over the state-of-the-art image classification approaches in automatic pavement distress detection. The source codes of IOPLIN are released on \url{https://github.com/DearCaat/ioplin}, and the CQU-BPDD dataset is able to be accessed on \url{https://dearcaat.github.io/CQU-BPDD/}.

cs.CV↗

arXiv2022

Recent Progress in Pencil Beam Scanning FLASH Proton Therapy: A Narrative Review

Background and Objective: Recent experimental studies using ultra-high dose rate radiation therapy (FLASH-RT) have shown improved normal tissue sparing and comparable tumor control compared to conventional dose rate RT. Pencil beam scanning (PBS) proton therapy with superior dosimetry characteristics has begun to draw attention to the delivery of conformal FLASH-RT for preclinical studies. This review aims to provide recent updates on the development of PBS FLASH-RT. Methods: The information summarized in this review article is based on search results in databases such as PubMed and search engines like Google Scholar, with keywords including pencil beam scanning, proton therapy, proton FLASH, Bragg peak FLASH, etc., with English articles from the year of 2014-2022. Content and Findings: This review summarizes of recent developments in PBS FLASH proton therapy (FLASH-PT), including PBS dose rate characterization, current delivery limitations, treatment planning, and biological investigations. Conclusions: As PBS FLASH delivery has enabled successful biological studies using transmission beams, the further improvement in PBS Bragg peak FLASH technologies will result in more advanced treatment plans associated with potentially improved outcomes.

physics.med-ph↗

arXiv2022

Boosting Multi-Label Image Classification with Complementary Parallel Self-Distillation

Multi-Label Image Classification (MLIC) approaches usually exploit label correlations to achieve good performance. However, emphasizing correlation like co-occurrence may overlook discriminative features of the target itself and lead to model overfitting, thus undermining the performance. In this study, we propose a generic framework named Parallel Self-Distillation (PSD) for boosting MLIC models. PSD decomposes the original MLIC task into several simpler MLIC sub-tasks via two elaborated complementary task decomposition strategies named Co-occurrence Graph Partition (CGP) and Dis-occurrence Graph Partition (DGP). Then, the MLIC models of fewer categories are trained with these sub-tasks in parallel for respectively learning the joint patterns and the category-specific patterns of labels. Finally, knowledge distillation is leveraged to learn a compact global ensemble of full categories with these learned patterns for reconciling the label correlation exploitation and model overfitting. Extensive results on MS-COCO and NUS-WIDE datasets demonstrate that our framework can be easily plugged into many MLIC approaches and improve performances of recent state-of-the-art approaches. The explainable visual study also further validates that our method is able to learn both the category-specific and co-occurring features. The source code is released at https://github.com/Robbie-Xu/CPSD.

cs.CV↗

arXiv2022

Higher-order topological pumping and its observation in photonic lattices

The discovery of the quantization of particle transport in adiabatic pumping cycles of periodic structures by Thouless [Phys. Rev. B 27, 6083 (1983)] linked the Chern number, a topological invariant characterizing the quantum Hall effect in two-dimensional electron gases, with the topology of dynamical periodic systems in one dimension. Here, we demonstrate its counterpart for higher-order topology. Specifically, we show that adiabatic cycles in two-dimensional crystals with vanishing dipole moments (and therefore zero overall particle transport) can nevertheless be topologically nontrivial. These cycles are associated with higher-order topology and can be diagnosed by their ability to produce corner-to-corner transport in certain metamaterial platforms. We experimentally verify the corner to corner transport associated with this topological pump by using an array of photonic waveguides adiabatically modulated in their separations and refractive indices. By mapping the dynamical phenomenon demonstrated here from two spatial and one temporal dimensions to three spatial dimensions, our observations are equivalent to the observation of chiral hinge states in a three-dimensional second-order topological insulator.

cond-mat.mes-hall↗

arXiv2021

Dizygotic Conditional Variational AutoEncoder for Multi-Modal and Partial Modality Absent Few-Shot Learning

Data augmentation is a powerful technique for improving the performance of the few-shot classification task. It generates more samples as supplements, and then this task can be transformed into a common supervised learning issue for solution. However, most mainstream data augmentation based approaches only consider the single modality information, which leads to the low diversity and quality of generated features. In this paper, we present a novel multi-modal data augmentation approach named Dizygotic Conditional Variational AutoEncoder (DCVAE) for addressing the aforementioned issue. DCVAE conducts feature synthesis via pairing two Conditional Variational AutoEncoders (CVAEs) with the same seed but different modality conditions in a dizygotic symbiosis manner. Subsequently, the generated features of two CVAEs are adaptively combined to yield the final feature, which can be converted back into its paired conditions while ensuring these conditions are consistent with the original conditions not only in representation but also in function. DCVAE essentially provides a new idea of data augmentation in various multi-modal scenarios by exploiting the complement of different modality prior information. Extensive experimental results demonstrate our work achieves state-of-the-art performances on miniImageNet, CIFAR-FS and CUB datasets, and is able to work well in the partial modality absence case.

cs.CV↗

arXiv2021

Deep Semantic Dictionary Learning for Multi-label Image Classification

Compared with single-label image classification, multi-label image classification is more practical and challenging. Some recent studies attempted to leverage the semantic information of categories for improving multi-label image classification performance. However, these semantic-based methods only take semantic information as type of complements for visual representation without further exploitation. In this paper, we present an innovative path towards the solution of the multi-label image classification which considers it as a dictionary learning task. A novel end-to-end model named Deep Semantic Dictionary Learning (DSDL) is designed. In DSDL, an auto-encoder is applied to generate the semantic dictionary from class-level semantics and then such dictionary is utilized for representing the visual features extracted by Convolutional Neural Network (CNN) with label embeddings. The DSDL provides a simple but elegant way to exploit and reconcile the label, semantic and visual spaces simultaneously via conducting the dictionary learning among them. Moreover, inspired by iterative optimization of traditional dictionary learning, we further devise a novel training strategy named Alternately Parameters Update Strategy (APUS) for optimizing DSDL, which alternately optimizes the representation coefficients and the semantic dictionary in forward and backward propagation. Extensive experimental results on three popular benchmarks demonstrate that our method achieves promising performances in comparison with the state-of-the-arts. Our codes and models have been released at {https://github.com/ZFT-CQU/DSDL}.

cs.CV↗

arXiv2021

Plot2API: Recommending Graphic API from Plot via Semantic Parsing Guided Neural Network

Plot-based Graphic API recommendation (Plot2API) is an unstudied but meaningful issue, which has several important applications in the context of software engineering and data visualization, such as the plotting guidance of the beginner, graphic API correlation analysis, and code conversion for plotting. Plot2API is a very challenging task, since each plot is often associated with multiple APIs and the appearances of the graphics drawn by the same API can be extremely varied due to the different settings of the parameters. Additionally, the samples of different APIs also suffer from extremely imbalanced. Considering the lack of technologies in Plot2API, we present a novel deep multi-task learning approach named Semantic Parsing Guided Neural Network (SPGNN) which translates the Plot2API issue as a multi-label image classification and an image semantic parsing tasks for the solution. In SPGNN, the recently advanced Convolutional Neural Network (CNN) named EfficientNet is employed as the backbone network for API recommendation. Meanwhile, a semantic parsing module is complemented to exploit the semantic relevant visual information in feature learning and eliminate the appearance-relevant visual information which may confuse the visual-information-based API recommendation. Moreover, the recent data augmentation technique named random erasing is also applied for alleviating the imbalance of API categories. We collect plots with the graphic APIs used to drawn them from Stack Overflow, and release three new Plot2API datasets corresponding to the graphic APIs of R and Python programming languages for evaluating the effectiveness of Plot2API techniques. Extensive experimental results not only demonstrate the superiority of our method over the recent deep learning baselines but also show the practicability of our method in the recommendation of graphic APIs.

cs.SE↗

arXiv2021

Oscars: Adaptive Semi-Synchronous Parallel Model for Distributed Deep Learning with Global View

Deep learning has become an indispensable part of life, such as face recognition, NLP, etc., but the training of deep model has always been a challenge, and in recent years, the complexity of training data and models has shown explosive growth, so the training method is gradually transformed into distributed training. Classical synchronization strategy can guarantee accuracy but frequent communication can lead to a slow training speed, although asynchronous strategy training speed but can not guarantee the accuracy, and in the face of the training of the heterogeneous cluster, the above work is not efficient to work, on the one hand, can cause serious waste of resources, on the other hand, frequent communication also made slow training speed, so this paper proposes a semi-synchronous training strategy based on local-SDG, effectively improve the utilization efficiency of heterogeneous resources cluster and reduce communication overhead, to accelerate the training and ensure the accuracy of the model.

cs.DC↗

arXiv2019

Braiding photonic topological zero modes

A remarkable property of quantum mechanics in two-dimensional (2D) space is its ability to support "anyons," particles that are neither fermions nor bosons. Theory predicts that these exotic excitations can be realized as bound states confined near topological defects, like Majorana zero modes trapped in vortices in topological superconductors. Intriguingly, in the simplest cases the nontrivial phase that arises when such defects are "braided" around one another is not intrinsically quantum mechanical; rather, it can be viewed as a manifestation of the geometric (Pancharatnam-Berry) phase in wave mechanics, enabling the simulation of such phenomena in classical systems. Here we report the first experimental measurement in any system, quantum or classical, of the geometric phase due to such a braiding process. These measurements are obtained using an interferometer constructed from highly tunable 2D arrays of photonic waveguides. Our results introduce photonic lattices as a versatile playground for the experimental study of topological defects and their braiding, complementing ongoing efforts in solid-state systems and cold atomic gases.

physics.optics↗

arXiv2018

Experimental realization of a Weyl exceptional ring

Weyl points are isolated degeneracies in reciprocal space that are monopoles of the Berry curvature. This topological charge makes them inherently robust to Hermitian perturbations of the system. However, non-Hermitian effects, usually inaccessible in condensed matter systems, are an important feature of photonics systems, and when added to an otherwise Hermitian Weyl material have been predicted to spread the Berry charge of the Weyl point out onto a ring of exceptional points, creating a Weyl exceptional ring and fundamentally altering its properties. Here, we observe the implications of the Weyl exceptional ring using real-space measurements of an evanescently-coupled bipartite optical waveguide array by probing its effects on the Fermi arc surface states, the bulk diffraction properties, and the output power ratio of the two constituent sublattices. This is the first realization of an object with topological Berry charge in a non-Hermitian system.

physics.optics↗

arXiv2017

Photonic Realization of a Transition Between Topological Phases Residing in the Weak and Strong Floquet Driving Regimes

A photonic Floquet topological insulator has previously been experimentally realized in an array of evanescently-coupled helical waveguides. In the topological regime probed by that experiment, the chirality of the single topological edge mode was the same as the chirality of the helices. Here we demonstrate the rich structure present in Floquet systems by fixing the helix chirality while moving into the strong driving regime to observe a topological transition in which the edge mode reverses its propagation direction, yielding the counterintuitive result that an increase of the driving amplitude can have the same effect as reversing the helix chirality. We experimentally observe this transition while overcoming the bending loss typically associated with the strong driving regime, despite the fact that the helix radius is on the same scale as the size of the entire array. The two topological phases observed in our experiment can be understood as originating from a Floquet realization of the Haldane model.

physics.optics↗

arXiv2017

Observation of Photonic Topological Valley-Hall Edge States

We experimentally demonstrate topological edge states arising from the valley-Hall effect in twodimensional honeycomb photonic lattices with broken inversion symmetry. We break inversion symmetry by detuning the refractive indices of the two honeycomb sublattices, giving rise to a boron nitride-like band structure. The edge states therefore exist along the domain walls between regions of opposite valley Chern numbers. We probe both the armchair and zig-zag domain walls and show that the former become gapped for any detuning, whereas the latter remain ungapped until a cutoff is reached. The valley-Hall effect provides a new mechanism for the realization of time-reversal invariant photonic topological insulators.

physics.optics↗

arXiv2017

Photonic topological pumping through the edges of a dynamical four-dimensional quantum Hall system

When a two-dimensional electron gas is exposed to a perpendicular magnetic field and an in-plane electric field, its conductance becomes quantized in the transverse in-plane direction: this is known as the quantum Hall (QH) effect. This effect is a result of the nontrivial topology of the system's electronic band structure, where an integer topological invariant known as the first Chern number leads to the quantization of the Hall conductance. Interestingly, it was shown that the QH effect can be generalized mathematically to four spatial dimensions (4D), but this effect has never been realized for the obvious reason that experimental systems are bound to three spatial dimensions. In this work, we harness the high tunability and control offered by photonic waveguide arrays to experimentally realize a dynamically-generated 4D QH system using a 2D array of coupled optical waveguides. The inter-waveguide separation is constructed such that the propagation of light along the device samples over higher-dimensional momenta in the directions orthogonal to the two physical dimensions, thus realizing a 2D topological pump. As a result, the device's band structure is associated with 4D topological invariants known as second Chern numbers which support a quantized bulk Hall response with a 4D symmetry. In a finite-sized system, the 4D topological bulk response is carried by localized edges modes that cross the sample as a function of of the modulated auxiliary momenta. We directly observe this crossing through photon pumping from edge-to-edge and corner-to-corner of our system. These are equivalent to the pumping of charge across a 4D system from one 3D hypersurface to the opposite one and from one 2D hyperedge to another, and serve as first experimental realization of higher-dimensional topological physics.

quant-ph↗

arXiv2016

Topological protection of photonic mid-gap cavity modes

Defect modes in two-dimensional periodic photonic structures have found use in a highly diverse set of optical devices. For example, photonic crystal cavities confine optical modes to subwavelength volumes and can be used for Purcell enhancement of nonlinearity, lasing, and cavity quantum electrodynamics. Photonic crystal fiber defect cores allow for supercontinuum generation and endlessly-single-mode fibers with large cores. However, these modes are notoriously fragile: small changes in the structure can lead to significant detuning of resonance frequency and mode volume. Here, we show that a photonic topological crystalline insulator structure can be used to topologically protect the resonance frequency to be in the middle of the band gap, and therefore minimize the mode volume of a two-dimensional photonic defect mode. We experimentally demonstrate this in a femtosecond-laser-written waveguide array, a geometry akin to a photonic crystal fiber. The topological defect modes are determined by a topological invariant that protects zero-dimensional states (defect modes) embedded in a two-dimensional environment; a novel form of topological protection that has not been previously demonstrated.

cond-mat.mes-hall↗

arXiv2016

Experimental observation of optical Weyl points

Weyl fermions are hypothetical two-component massless relativistic particles in three-dimensional (3D) space, proposed by Hermann Weyl in 1929. Their band-crossing points, called 'Weyl points,' carry a topological charge and are therefore highly robust. There has been much excitement over recent observations of Weyl points in microwave photonic crystals and the semimetal TaAs. Here, we report on the first experimental observation of Weyl points of light at optical frequencies. These are also the first observations of 'type-II' Weyl points for photons, which have strictly positive group velocity along one spatial direction. We use a 3D structure consisting of laser-written waveguides, and show the presence of type-II Weyl points by (1) observing conical diffraction along one axis when the frequency is tuned to the Weyl point; and (2) observing the associated Fermi arc surface states. The realization of Weyl points at optical frequencies allow these novel electromagnetic modes to be further explored in the context of linear, nonlinear, and quantum optics.

physics.optics↗

arXiv2016

Regression-based Hypergraph Learning for Image Clustering and Classification

Inspired by the recently remarkable successes of Sparse Representation (SR), Collaborative Representation (CR) and sparse graph, we present a novel hypergraph model named Regression-based Hypergraph (RH) which utilizes the regression models to construct the high quality hypergraphs. Moreover, we plug RH into two conventional hypergraph learning frameworks, namely hypergraph spectral clustering and hypergraph transduction, to present Regression-based Hypergraph Spectral Clustering (RHSC) and Regression-based Hypergraph Transduction (RHT) models for addressing the image clustering and classification issues. Sparse Representation and Collaborative Representation are employed to instantiate two RH instances and their RHSC and RHT algorithms. The experimental results on six popular image databases demonstrate that the proposed RH learning algorithms achieve promising image clustering and classification performances, and also validate that RH can inherit the desirable properties from both hypergraph models and regression models.

cs.CV↗

arXiv2015

Beam specific planning target volumes incorporating 4DCT for pencil beam scanning proton therapy of thoracic tumors

The purpose of this study is to determine whether organ sparing and target coverage can be simultaneously maintained for pencil beam scanning (PBS) proton therapy treatment of thoracic tumors in the presence of motion, stopping power uncertainties and patient setup variations. Ten consecutive patients that were previously treated with proton therapy to 66.6/1.8 Gy (RBE) using double scattering (DS) were replanned with PBS. Minimum and maximum intensity images from 4DCT were used to introduce flexible smearing in the determination of the beam specific PTV (BSPTV). Datasets from eight 4DCT phases, using +-3% uncertainty in stopping power, and +-3 mm uncertainty in patient setup in each direction were used to create 8X12X10=960 PBS plans for the evaluation of ten patients. Plans were normalized to provide identical coverage between DS and PBS. The average lung V20, V5, and mean doses were reduced from 29.0%, 35.0%, and 16.4 Gy with DS to 24.6%, 30.6%, and 14.1 Gy with PBS, respectively. The average heart V30 and V45 were reduced from 10.4% and 7.5% in DS to 8.1% and 5.4% for PBS, respectively. Furthermore, the maximum spinal cord, esophagus and heart dose were decreased from 37.1 Gy, 71.7 Gy and 69.2 Gy with DS to 31.3 Gy, 67.9 Gy and 64.6 Gy with PBS. The conformity index (CI), homogeneity index (HI), and global maximal dose were improved from 3.2, 0.08, 77.4 Gy with DS to 2.8, 0.04 and 72.1 Gy with PBS. All differences are statistically significant, with p values <0.05, with the exception of the heart V45 (p= 0.146). PBS with BSPTV achieves better organ sparing and improves target coverage using a repainting method for the treatment of thoracic tumors. Incorporating motion-related uncertainties is essential in maintaining marginal coverage and homogenous dose of treatment targets.

physics.med-ph↗

arXiv2015

Learning Hypergraph-regularized Attribute Predictors

We present a novel attribute learning framework named Hypergraph-based Attribute Predictor (HAP). In HAP, a hypergraph is leveraged to depict the attribute relations in the data. Then the attribute prediction problem is casted as a regularized hypergraph cut problem in which HAP jointly learns a collection of attribute projections from the feature space to a hypergraph embedding space aligned with the attribute space. The learned projections directly act as attribute classifiers (linear and kernelized). This formulation leads to a very efficient approach. By considering our model as a multi-graph cut task, our framework can flexibly incorporate other available information, in particular class label. We apply our approach to attribute prediction, Zero-shot and $N$-shot learning tasks. The results on AWA, USAA and CUB databases demonstrate the value of our methods in comparison with the state-of-the-art approaches.

cs.CV↗