arXiv subjects

Chao Dong

Publications and source records attributed to Chao Dong.

At least 181 records · Page 10Linked to original sources

arXiv2019

Identification of Light Sources using Machine Learning

The identification of light sources represents a task of utmost importance for the development of multiple photonic technologies. Over the last decades, the identification of light sources as diverse as sunlight, laser radiation and molecule fluorescence has relied on the collection of photon statistics or the implementation of quantum state tomography. In general, this task requires an extensive number of measurements to unveil the characteristic statistical fluctuations and correlation properties of light, particularly in the low-photon flux regime. In this article, we exploit the self-learning features of artificial neural networks and naive Bayes classifier to dramatically reduce the number of measurements required to discriminate thermal light from coherent light at the single-photon level. We demonstrate robust light identification with tens of measurements at mean photon numbers below one. Our work demonstrates an improvement in terms of the number of measurements of several orders of magnitude with respect to conventional schemes for characterization of light sources. Our work has important implications for multiple photonic technologies such as LIDAR and microscopy.

quant-ph↗

arXiv2019

Approaching the Finite Blocklength Capacity within 0.025dB by Short Polar Codes and CRC-Aided Hybrid Decoding

In this letter, we explore the performance limits of short polar codes and find that the maximum likelihood (ML) performance of a simple CRC-polar concatenated scheme can approach the finite blocklength capacity. Then, in order to approach the ML performance with a low average complexity, a CRC-aided hybrid decoding (CA-HD) algorithm is proposed and its decoding process is divided into two steps. In the first step, the received sequence is decoded by the adaptive successive cancellation list (ADSCL) decoding. In the second step, CRC-aided sphere decoding with a reasonable initial radius is used to decode the received sequence. To obtain the reasonable radius, the CRC bits of the survival paths in ADSCL are recalculated and the minimum Euclidean distance between the survival path and the received sequence is chosen as the initial radius. The simulation results show that CA-HD can achieve within about $0.025$dB of the finite blocklength capacity at the block error ratio $10^{-3}$ with code length $128$ and code rate $1/2$.

cs.IT↗

arXiv2019

RankSRGAN: Generative Adversarial Networks with Ranker for Image Super-Resolution

Generative Adversarial Networks (GAN) have demonstrated the potential to recover realistic details for single image super-resolution (SISR). To further improve the visual quality of super-resolved results, PIRM2018-SR Challenge employed perceptual metrics to assess the perceptual quality, such as PI, NIQE, and Ma. However, existing methods cannot directly optimize these indifferentiable perceptual metrics, which are shown to be highly correlated with human ratings. To address the problem, we propose Super-Resolution Generative Adversarial Networks with Ranker (RankSRGAN) to optimize generator in the direction of perceptual metrics. Specifically, we first train a Ranker which can learn the behavior of perceptual metrics and then introduce a novel rank-content loss to optimize the perceptual quality. The most appealing part is that the proposed method can combine the strengths of different SR methods to generate better results. Extensive experiments show that RankSRGAN achieves visually pleasing results and reaches state-of-the-art performance in perceptual metrics. Project page: https://wenlongzhang0724.github.io/Projects/RankSRGAN

cs.CV↗

arXiv2019

Thermodynamic Investigation of Metamagnetism in Pulsed High Magnetic Fields on Heavy Fermion Superconductor UTe$_2$

We investigated the thermodynamic property of the heavy fermion superconductor UTe$_2$ in pulsed high magnetic fields. The superconducting transition in zero field was observed at $T_{\rm c}$=1.65 K as a sharp heat capacity jump. Magnetocaloric effect measurements in pulsed-magnetic fields obviously detected a thermodynamic anomaly accompanied by a first-order metamagnetic transition at $μ$$_{0}$$H_{\rm m}$=36.0 T when the fields are applied nearly along the hard-magnetization $b$-axis. From the results of heat capacity measurements in magnetic fields, we found a drastic diverging electronic heat capacity coefficient of the normal state $γ$$_{\rm N}$ with approaching $H_{\rm m}$. Comparing with the previous works via the magnetic Clausius-Clapeyron relation, we unveil the thermodynamic details of the metamagnetic transition. The enhancement of the effective mass observed as the development of $γ_{\rm N}$ indicates that quantum fluctuation strongly evolves around $H_{\rm m}$; it assists the superconductivity emerging even in extremely high fields.

cond-mat.str-el↗

arXiv2019

Suppressing Model Overfitting for Image Super-Resolution Networks

Large deep networks have demonstrated competitive performance in single image super-resolution (SISR), with a huge volume of data involved. However, in real-world scenarios, due to the limited accessible training pairs, large models exhibit undesirable behaviors such as overfitting and memorization. To suppress model overfitting and further enjoy the merits of large model capacity, we thoroughly investigate generic approaches for supplying additional training data pairs. In particular, we introduce a simple learning principle MixUp to train networks on interpolations of sample pairs, which encourages networks to support linear behavior in-between training samples. In addition, we propose a data synthesis method with learned degradation, enabling models to use extra high-quality images with higher content diversity. This strategy proves to be successful in reducing biases of data. By combining these components -- MixUp and synthetic training data, large models can be trained without overfitting under very limited data samples and achieve satisfactory generalization performance. Our method won the second place in NTIRE2019 Real SR Challenge.

cs.CV↗

arXiv2019

Modulating Image Restoration with Continual Levels via Adaptive Feature Modification Layers

In image restoration tasks, like denoising and super resolution, continual modulation of restoration levels is of great importance for real-world applications, but has failed most of existing deep learning based image restoration methods. Learning from discrete and fixed restoration levels, deep models cannot be easily generalized to data of continuous and unseen levels. This topic is rarely touched in literature, due to the difficulty of modulating well-trained models with certain hyper-parameters. We make a step forward by proposing a unified CNN framework that consists of few additional parameters than a single-level model yet could handle arbitrary restoration levels between a start and an end level. The additional module, namely AdaFM layer, performs channel-wise feature modification, and can adapt a model to another restoration level with high accuracy. By simply tweaking an interpolation coefficient, the intermediate model - AdaFM-Net could generate smooth and continuous restoration effects without artifacts. Extensive experiments on three image restoration tasks demonstrate the effectiveness of both model training and modulation testing. Besides, we carefully investigate the properties of AdaFM layers, providing a detailed guidance on the usage of the proposed method.

cs.CV↗

arXiv2019

Blind Super-Resolution With Iterative Kernel Correction

Deep learning based methods have dominated super-resolution (SR) field due to their remarkable performance in terms of effectiveness and efficiency. Most of these methods assume that the blur kernel during downsampling is predefined/known (e.g., bicubic). However, the blur kernels involved in real applications are complicated and unknown, resulting in severe performance drop for the advanced SR methods. In this paper, we propose an Iterative Kernel Correction (IKC) method for blur kernel estimation in blind SR problem, where the blur kernels are unknown. We draw the observation that kernel mismatch could bring regular artifacts (either over-sharpening or over-smoothing), which can be applied to correct inaccurate blur kernels. Thus we introduce an iterative correction scheme -- IKC that achieves better results than direct kernel estimation. We further propose an effective SR network architecture using spatial feature transform (SFT) layers to handle multiple blur kernels, named SFTMD. Extensive experiments on synthetic and real-world images show that the proposed IKC method with SFTMD can provide visually favorable SR results and the state-of-the-art performance in blind SR problem.

cs.CV↗

arXiv2019

EDVR: Video Restoration with Enhanced Deformable Convolutional Networks

Video restoration tasks, including super-resolution, deblurring, etc, are drawing increasing attention in the computer vision community. A challenging benchmark named REDS is released in the NTIRE19 Challenge. This new benchmark challenges existing methods from two aspects: (1) how to align multiple frames given large motions, and (2) how to effectively fuse different frames with diverse motion and blur. In this work, we propose a novel Video Restoration framework with Enhanced Deformable networks, termed EDVR, to address these challenges. First, to handle large motions, we devise a Pyramid, Cascading and Deformable (PCD) alignment module, in which frame alignment is done at the feature level using deformable convolutions in a coarse-to-fine manner. Second, we propose a Temporal and Spatial Attention (TSA) fusion module, in which attention is applied both temporally and spatially, so as to emphasize important features for subsequent restoration. Thanks to these modules, our EDVR wins the champions and outperforms the second place by a large margin in all four tracks in the NTIRE19 video restoration and enhancement challenges. EDVR also demonstrates superior performance to state-of-the-art published methods on video super-resolution and deblurring. The code is available at https://github.com/xinntao/EDVR.

cs.CV↗

arXiv2019

Identifying the Best Machine Learning Algorithms for Brain Tumor Segmentation, Progression Assessment, and Overall Survival Prediction in the BRATS Challenge

Gliomas are the most common primary brain malignancies, with different degrees of aggressiveness, variable prognosis and various heterogeneous histologic sub-regions, i.e., peritumoral edematous/invaded tissue, necrotic core, active and non-enhancing core. This intrinsic heterogeneity is also portrayed in their radio-phenotype, as their sub-regions are depicted by varying intensity profiles disseminated across multi-parametric magnetic resonance imaging (mpMRI) scans, reflecting varying biological properties. Their heterogeneous shape, extent, and location are some of the factors that make these tumors difficult to resect, and in some cases inoperable. The amount of resected tumor is a factor also considered in longitudinal scans, when evaluating the apparent tumor for potential diagnosis of progression. Furthermore, there is mounting evidence that accurate segmentation of the various tumor sub-regions can offer the basis for quantitative image analysis towards prediction of patient overall survival. This study assesses the state-of-the-art machine learning (ML) methods used for brain tumor image analysis in mpMRI scans, during the last seven instances of the International Brain Tumor Segmentation (BraTS) challenge, i.e., 2012-2018. Specifically, we focus on i) evaluating segmentations of the various glioma sub-regions in pre-operative mpMRI scans, ii) assessing potential tumor progression by virtue of longitudinal growth of tumor sub-regions, beyond use of the RECIST/RANO criteria, and iii) predicting the overall survival from pre-operative mpMRI scans of patients that underwent gross total resection. Finally, we investigate the challenge of identifying the best ML algorithms for each of these tasks, considering that apart from being diverse on each instance of the challenge, the multi-institutional mpMRI BraTS dataset has also been a continuously evolving/growing dataset.

cs.CV↗

arXiv2019

Quantum Critical Dynamics of a Heisenberg-Ising Chain in a Longitudinal Field: Many-Body Strings versus Fractional Excitations

We report a high-resolution terahertz spectroscopic study of quantum spin dynamics in the antiferromagnetic Heisenberg-Ising spin-chain compound BaCo$_2$V$_2$O$_8$ as a function of temperature and longitudinal magnetic field. Confined spinon excitations are observed in an antiferromagnetic phase below $T_N\simeq 5.5$ K. In a field-induced gapless phase above $B_c=3.8$ T, we identify many-body string excitations as well as low-energy fractional psinon/antipsinon excitations by comparing to Bethe-Ansatz calculations. In the vicinity of $B_c$, the high-energy string excitations are found to be dynamically dominant over the fractional excitations.

cond-mat.str-el↗

arXiv2019

Field-induced double dome and Bose-Einstein condensation in the crossing quantum spin chain system AgVOAsO4

We present inelastic neutron scattering data on the quantum paramagnet AgVOAsO4 that establish the system is a S=1/2 alternating spin chain compound and provide a direct measurement of the spin gap. We also present experimental evidence for two different types of field-induced magnetic order between mu_0H_c1 = 8.4T and mu_0H_c2 = 48.9T, which may be related to Bose-Einstein condensation (BEC) of triplons. Thermodynamic measurements in magnetic fields up to 60T and temperatures down to 0.1K reveal a H-T phase diagram consisting of a dome encapsulating two ordered phases with maximum ordering temperatures of 3.8K and 5.3K respectively. This complex phase diagram is not expected for a single-Q BEC system and therefore establishes AgVOAsO4 as a promising multi-Q BEC candidate capable of hosting exotic vortex phases.

cond-mat.str-el↗

arXiv2018

Deep Network Interpolation for Continuous Imagery Effect Transition

Deep convolutional neural network has demonstrated its capability of learning a deterministic mapping for the desired imagery effect. However, the large variety of user flavors motivates the possibility of continuous transition among different output effects. Unlike existing methods that require a specific design to achieve one particular transition (e.g., style transfer), we propose a simple yet universal approach to attain a smooth control of diverse imagery effects in many low-level vision tasks, including image restoration, image-to-image translation, and style transfer. Specifically, our method, namely Deep Network Interpolation (DNI), applies linear interpolation in the parameter space of two or more correlated networks. A smooth control of imagery effects can be achieved by tweaking the interpolation coefficients. In addition to DNI and its broad applications, we also investigate the mechanism of network interpolation from the perspective of learned filters.

cs.CV↗

arXiv2018

Hierarchical multi-class segmentation of glioma images using networks with multi-level activation function

For many segmentation tasks, especially for the biomedical image, the topological prior is vital information which is useful to exploit. The containment/nesting is a typical inter-class geometric relationship. In the MICCAI Brain tumor segmentation challenge, with its three hierarchically nested classes 'whole tumor', 'tumor core', 'active tumor', the nested classes relationship is introduced into the 3D-residual-Unet architecture. The network comprises a context aggregation pathway and a localization pathway, which encodes increasingly abstract representation of the input as going deeper into the network, and then recombines these representations with shallower features to precisely localize the interest domain via a localization path. The nested-class-prior is combined by proposing the multi-class activation function and its corresponding loss function. The model is trained on the training dataset of Brats2018, and 20% of the dataset is regarded as the validation dataset to determine parameters. When the parameters are fixed, we retrain the model on the whole training dataset. The performance achieved on the validation leaderboard is 86%, 77% and 72% Dice scores for the whole tumor, enhancing tumor and tumor core classes without relying on ensembles or complicated post-processing steps. Based on the same start-of-the-art network architecture, the accuracy of nested-class (enhancing tumor) is reasonably improved from 69% to 72% compared with the traditional Softmax-based method which blind to topological prior.

cs.CV↗

arXiv2018

PIRM Challenge on Perceptual Image Enhancement on Smartphones: Report

This paper reviews the first challenge on efficient perceptual image enhancement with the focus on deploying deep learning models on smartphones. The challenge consisted of two tracks. In the first one, participants were solving the classical image super-resolution problem with a bicubic downscaling factor of 4. The second track was aimed at real-world photo enhancement, and the goal was to map low-quality photos from the iPhone 3GS device to the same photos captured with a DSLR camera. The target metric used in this challenge combined the runtime, PSNR scores and solutions' perceptual results measured in the user study. To ensure the efficiency of the submitted models, we additionally measured their runtime and memory requirements on Android smartphones. The proposed solutions significantly improved baseline results defining the state-of-the-art for image enhancement on smartphones.

cs.CV↗

arXiv2018

ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks

The Super-Resolution Generative Adversarial Network (SRGAN) is a seminal work that is capable of generating realistic textures during single image super-resolution. However, the hallucinated details are often accompanied with unpleasant artifacts. To further enhance the visual quality, we thoroughly study three key components of SRGAN - network architecture, adversarial loss and perceptual loss, and improve each of them to derive an Enhanced SRGAN (ESRGAN). In particular, we introduce the Residual-in-Residual Dense Block (RRDB) without batch normalization as the basic network building unit. Moreover, we borrow the idea from relativistic GAN to let the discriminator predict relative realness instead of the absolute value. Finally, we improve the perceptual loss by using the features before activation, which could provide stronger supervision for brightness consistency and texture recovery. Benefiting from these improvements, the proposed ESRGAN achieves consistently better visual quality with more realistic and natural textures than SRGAN and won the first place in the PIRM2018-SR Challenge. The code is available at https://github.com/xinntao/ESRGAN .

cs.CV↗

arXiv2018

Unsupervised Image Super-Resolution using Cycle-in-Cycle Generative Adversarial Networks

We consider the single image super-resolution problem in a more general case that the low-/high-resolution pairs and the down-sampling process are unavailable. Different from traditional super-resolution formulation, the low-resolution input is further degraded by noises and blurring. This complicated setting makes supervised learning and accurate kernel estimation impossible. To solve this problem, we resort to unsupervised learning without paired data, inspired by the recent successful image-to-image translation applications. With generative adversarial networks (GAN) as the basic component, we propose a Cycle-in-Cycle network structure to tackle the problem within three steps. First, the noisy and blurry input is mapped to a noise-free low-resolution space. Then the intermediate image is up-sampled with a pre-trained deep model. Finally, we fine-tune the two modules in an end-to-end manner to get the high-resolution output. Experiments on NTIRE2018 datasets demonstrate that the proposed unsupervised method achieves comparable results as the state-of-the-art supervised models.

cs.CV↗

arXiv2018

Crafting a Toolchain for Image Restoration by Deep Reinforcement Learning

We investigate a novel approach for image restoration by reinforcement learning. Unlike existing studies that mostly train a single large network for a specialized task, we prepare a toolbox consisting of small-scale convolutional networks of different complexities and specialized in different tasks. Our method, RL-Restore, then learns a policy to select appropriate tools from the toolbox to progressively restore the quality of a corrupted image. We formulate a step-wise reward function proportional to how well the image is restored at each step to learn the action policy. We also devise a joint learning scheme to train the agent and tools for better performance in handling uncertainty. In comparison to conventional human-designed networks, RL-Restore is capable of restoring images corrupted with complex and unknown distortions in a more parameter-efficient manner using the dynamically formed toolchain.

cs.CV↗

arXiv2018

Recovering Realistic Texture in Image Super-resolution by Deep Spatial Feature Transform

Despite that convolutional neural networks (CNN) have recently demonstrated high-quality reconstruction for single-image super-resolution (SR), recovering natural and realistic texture remains a challenging problem. In this paper, we show that it is possible to recover textures faithful to semantic classes. In particular, we only need to modulate features of a few intermediate layers in a single network conditioned on semantic segmentation probability maps. This is made possible through a novel Spatial Feature Transform (SFT) layer that generates affine transformation parameters for spatial-wise feature modulation. SFT layers can be trained end-to-end together with the SR network using the same loss function. During testing, it accepts an input image of arbitrary size and generates a high-resolution image with just a single forward pass conditioned on the categorical priors. Our final results show that an SR network equipped with SFT can generate more realistic and visually pleasing textures in comparison to state-of-the-art SRGAN and EnhanceNet.

cs.CV↗