arXiv subjects

Feng Wu

Publications and source records attributed to Feng Wu.

At least 253 records · Page 14Linked to original sources

arXiv2018

Context-Aware Visual Policy Network for Sequence-Level Image Captioning

Many vision-language tasks can be reduced to the problem of sequence prediction for natural language output. In particular, recent advances in image captioning use deep reinforcement learning (RL) to alleviate the "exposure bias" during training: ground-truth subsequence is exposed in every step prediction, which introduces bias in test when only predicted subsequence is seen. However, existing RL-based image captioning methods only focus on the language policy while not the visual policy (e.g., visual attention), and thus fail to capture the visual context that are crucial for compositional reasoning such as visual relationships (e.g., "man riding horse") and comparisons (e.g., "smaller cat"). To fill the gap, we propose a Context-Aware Visual Policy network (CAVP) for sequence-level image captioning. At every time step, CAVP explicitly accounts for the previous visual attentions as the context, and then decides whether the context is helpful for the current word generation given the current visual attention. Compared against traditional visual attention that only fixes a single image region at every step, CAVP can attend to complex visual compositions over time. The whole image captioning model --- CAVP and its subsequent language policy network --- can be efficiently optimized end-to-end by using an actor-critic policy gradient method with respect to any caption evaluation metric. We demonstrate the effectiveness of CAVP by state-of-the-art performances on MS-COCO offline split and online server, using various metrics and sensible visualizations of qualitative visual context. The code is available at https://github.com/daqingliu/CAVP

cs.CV↗

arXiv2018

Significant enhancement of magneto-optical effect in one-dimensional photonic crystals with magnetized epsilon-near-zero defect

Nonreciprocal (NOR) transmission with magneto-optical materials plays a critical role in a broad range of applications, such as optical isolation, all-optical signal processing, and integrated photonic circuits. The underlying mechanism is that a static magnetic field can break the time-reversal symmetry in the presence of magneto-optical materials. However, the typical NOR devices usually need a large size because the weak magneto-optical activity of materials. Here, our theoretical investigations show that the NOR transmission can be obtained in the one-dimensional photonic crystal with a magnetized epsilon-near-zero (ENZ) defect due to strong field localization in the ENZ medium. The inherent weak magneto-optical activity is significantly enhanced in ENZ medium. In our configuration, the wavelength shift of transmission peaks along two opposite incident directions can be up to 100 times higher than that in the case that the defect layer is a normal bismuth iron garnet. Such results will provide a new degree to design novel magneto-optical devices with small size, and may open novel routes to exploit advanced materials for steering the electromagnetic waves in nano-scale structures.

physics.optics↗

arXiv2017

Note on Weyl versus Conformal Invariance in Field Theory

It was argued recently that conformal invariance in flat spacetime implies Weyl invariance in a general curved background for unitary theories and possible anomalies in the Weyl variation of scalar operators are identified. We argue that generically unitarity alone is not sufficient for a conformal field theory to be Weyl invariant. Furthermore, we show explicitly that when a unitary conformal field theory couples to gravity in a Weyl invariant way, each primary scalar operator that is either relevant or marginal in the unitary conformal field theory corresponds to a Weyl-covariant operator in the curved background.

hep-th↗

arXiv2017

Effects of tensor and $T=0$ pairing interactions on nuclear $β^+$/EC decay

The Hartree-Fock-Bogolyubov (HFB) plus proton-neutron quasiparticle random phase approximation (pnQRPA) approach based on Skyrme interaction is applied to study the nuclear $β^+$/EC decay for nuclei near the proton magic numbers $Z=$20, 28, and 50. With properly selected Skyrme interactions that include the tensor terms, and the $T=0$ pairing interaction, the experimental $β^+$/EC decay half-lives of these nuclei can be systematically reproduced quite well. It is shown that the tensor and $T=0$ pairing interactions play different roles in different nuclei. The specific effect is relevant to the configurations that contribute to the decay. The attractive and repulsive properties of tensor interaction for the GT state is also studied. The present results indicate that it works repulsively on the GT states dominated by the configurations from $πj_\gtrless$ to $νj_\gtrless$ orbits, while working attractively for those mainly composed of the configurations from $πj_\gtrless$ to $νj_\lessgtr$ orbits.

nucl-th↗

arXiv2017

First-principles Engineering of Charged Defects for Two-dimensional Quantum Technologies

Charged defects in 2D materials have emerging applications in quantum technologies such as quantum emitters and quantum computation. Advancement of these technologies requires rational design of ideal defect centers, demanding reliable computation methods for quantitatively accurate prediction of defect properties. We present an accurate, parameter-free and efficient procedure to evaluate quasiparticle defect states and thermodynamic charge transition levels of defects in 2D materials. Importantly, we solve critical issues that stem from the strongly anisotropic screening in 2D materials, that have so far precluded accurate prediction of charge transition levels in these materials. Using this procedure, we investigate various defects in monolayer hexagonal boron nitride (h-BN) for their charge transition levels, stable spin states and optical excitations. We identify $C_BN_V$ (nitrogen vacancy adjacent to carbon substitution of boron) to be the most promising defect candidate for scalable quantum bit and emitter applications.

cond-mat.mtrl-sci↗

arXiv2017

Neural network-based arithmetic coding of intra prediction modes in HEVC

In both H.264 and HEVC, context-adaptive binary arithmetic coding (CABAC) is adopted as the entropy coding method. CABAC relies on manually designed binarization processes as well as handcrafted context models, which may restrict the compression efficiency. In this paper, we propose an arithmetic coding strategy by training neural networks, and make preliminary studies on coding of the intra prediction modes in HEVC. Instead of binarization, we propose to directly estimate the probability distribution of the 35 intra prediction modes with the adoption of a multi-level arithmetic codec. Instead of handcrafted context models, we utilize convolutional neural network (CNN) to perform the probability estimation. Simulation results show that our proposed arithmetic coding leads to as high as 9.9% bits saving compared with CABAC.

cs.MM↗

arXiv2017

An Iterative BP-CNN Architecture for Channel Decoding

Inspired by recent advances in deep learning, we propose a novel iterative BP-CNN architecture for channel decoding under correlated noise. This architecture concatenates a trained convolutional neural network (CNN) with a standard belief-propagation (BP) decoder. The standard BP decoder is used to estimate the coded bits, followed by a CNN to remove the estimation errors of the BP decoder and obtain a more accurate estimation of the channel noise. Iterating between BP and CNN will gradually improve the decoding SNR and hence result in better decoding performance. To train a well-behaved CNN model, we define a new loss function which involves not only the accuracy of the noise estimation but also the normality test for the estimation errors, i.e., to measure how likely the estimation errors follow a Gaussian distribution. The introduction of the normality test to the CNN training shapes the residual noise distribution and further reduces the BER of the iterative decoding, compared to using the standard quadratic loss function. We carry out extensive experiments to analyze and verify the proposed framework. The iterative BP-CNN decoder has better BER performance with lower complexity, is suitable for parallel implementation, does not rely on any specific channel model or encoding method, and is robust against training mismatches. All of these features make it a good candidate for decoding modern channel codes.

stat.ML↗

arXiv2017

Convolutional Neural Network-Based Block Up-sampling for Intra Frame Coding

Inspired by the recent advances of image super-resolution using convolutional neural network (CNN), we propose a CNN-based block up-sampling scheme for intra frame coding. A block can be down-sampled before being compressed by normal intra coding, and then up-sampled to its original resolution. Different from previous studies on down/up-sampling-based coding, the up-sampling methods in our scheme have been designed by training CNN instead of hand-crafted. We explore a new CNN structure for up-sampling, which features deconvolution of feature maps, multi-scale fusion, and residue learning, making the network both compact and efficient. We also design different networks for the up-sampling of luma and chroma components, respectively, where the chroma up-sampling CNN utilizes the luma information to boost its performance. In addition, we design a two-stage up-sampling process, the first stage being within the block-by-block coding loop, and the second stage being performed on the entire frame, so as to refine block boundaries. We also empirically study how to set the coding parameters of down-sampled blocks for pursuing the frame-level rate-distortion optimization. Our proposed scheme is implemented into the High Efficiency Video Coding (HEVC) reference software, and a comprehensive set of experiments have been performed to evaluate our methods. Experimental results show that our scheme achieves significant bits saving compared with HEVC anchor especially at low bit rates, leading to on average 5.5% BD-rate reduction on common test sequences and on average 9.0% BD-rate reduction on ultra high definition (UHD) test sequences.

cs.MM↗

arXiv2017

Existence problem of proton semi-bubble structure in the $2_1^+$ state of $^{34}$Si

The fully self-consistent Hartree-Fock (HF) plus random phase approximation (RPA) based on Skyrme-type interaction is used to study the existence problem of proton semi-bubble structure in the $2_1^+$ state of $^{34}$Si. The experimental excitation energy and the B(E2) strength of the $2_1^+$ state in $^{34}$Si can be reproduced quite well. The tensor effect is also studied. It is shown that the tensor interaction has a notable impact on the excitation energy of the $2_1^+$ state and a small effect on the B(E2) value. Besides, its effect on the density distributions in the ground and $2_1^+$ state of $^{34}$Si is negligible. Our present results with T36 and T44 show that the $2_1^+$ state of $^{34}$Si is mainly caused by proton transiton from $π1d_{5/2}$ orbit to $π2s_{1/2}$ orbit, and the existence of a proton semi-bubble structure in this state is very unlikely.

nucl-th↗

arXiv2017

A Convolutional Neural Network Approach for Half-Pel Interpolation in Video Coding

Motion compensation is a fundamental technology in video coding to remove the temporal redundancy between video frames. To further improve the coding efficiency, sub-pel motion compensation has been utilized, which requires interpolation of fractional samples. The video coding standards usually adopt fixed interpolation filters that are derived from the signal processing theory. However, as video signal is not stationary, the fixed interpolation filters may turn out less efficient. Inspired by the great success of convolutional neural network (CNN) in computer vision, we propose to design a CNN-based interpolation filter (CNNIF) for video coding. Different from previous studies, one difficulty for training CNNIF is the lack of ground-truth since the fractional samples are actually not available. Our solution for this problem is to derive the "ground-truth" of fractional samples by smoothing high-resolution images, which is verified to be effective by the conducted experiments. Compared to the fixed half-pel interpolation filter for luma in High Efficiency Video Coding (HEVC), our proposed CNNIF achieves up to 3.2% and on average 0.9% BD-rate reduction under low-delay P configuration.

cs.MM↗

arXiv2017

An Efficient Four-Parameter Affine Motion Model for Video Coding

In this paper, we study a simplified affine motion model based coding framework to overcome the limitation of translational motion model and maintain low computational complexity. The proposed framework mainly has three key contributions. First, we propose to reduce the number of affine motion parameters from 6 to 4. The proposed four-parameter affine motion model can not only handle most of the complex motions in natural videos but also save the bits for two parameters. Second, to efficiently encode the affine motion parameters, we propose two motion prediction modes, i.e., advanced affine motion vector prediction combined with a gradient-based fast affine motion estimation algorithm and affine model merge, where the latter attempts to reuse the affine motion parameters (instead of the motion vectors) of neighboring blocks. Third, we propose two fast affine motion compensation algorithms. One is the one-step sub-pixel interpolation, which reduces the computations of each interpolation. The other is the interpolation-precision-based adaptive block size motion compensation, which performs motion compensation at the block level rather than the pixel level to reduce the interpolation times. Our proposed techniques have been implemented based on the state-of-the-art high efficiency video coding standard, and the experimental results show that the proposed techniques altogether achieve on average 11.1% and 19.3% bits saving for random access and low delay configurations, respectively, on typical video sequences that have rich rotation or zooming motions. Meanwhile, the computational complexity increases of both encoder and decoder are within an acceptable range.

cs.MM↗

arXiv2016

A Convolutional Neural Network Approach for Post-Processing in HEVC Intra Coding

Lossy image and video compression algorithms yield visually annoying artifacts including blocking, blurring, and ringing, especially at low bit-rates. To reduce these artifacts, post-processing techniques have been extensively studied. Recently, inspired by the great success of convolutional neural network (CNN) in computer vision, some researches were performed on adopting CNN in post-processing, mostly for JPEG compressed images. In this paper, we present a CNN-based post-processing algorithm for High Efficiency Video Coding (HEVC), the state-of-the-art video coding standard. We redesign a Variable-filter-size Residue-learning CNN (VRCNN) to improve the performance and to accelerate network training. Experimental results show that using our VRCNN as post-processing leads to on average 4.6% bit-rate reduction compared to HEVC baseline. The VRCNN outperforms previously studied networks in achieving higher bit-rate reduction, lower memory cost, and multiplied computational speedup.

cs.MM↗

arXiv2016

Aspects of a nonminimal conformal extension of the standard model

In this article we investigate a conformal extension of the standard model in which the scalar sector consists of a standard model Higgs doublet, a real gauge singlet and a real $SU(2)_{L}$ triplet. Focusing on the scenario where the Higgs boson found at the LHC is identified as the pseudo-Nambu-Goldstone boson of broken scale invariance, various theoretical and phenomenological features of the model are discussed. In particular, we analyze the decay pattern of the new scalar resonance. We also show that when the mass of this new scalar resonance is far below the $WW$ threshold, the natural regions of the parameter space are reduced by a constraint associated with the symmetry enhancement due to the decoupling of the singlet scalar from the rest of the system.

hep-ph↗

arXiv2016

Understanding User Instructions by Utilizing Open Knowledge for Service Robots

Understanding user instructions in natural language is an active research topic in AI and robotics. Typically, natural user instructions are high-level and can be reduced into low-level tasks expressed in common verbs (e.g., `take', `get', `put'). For robots understanding such instructions, one of the key challenges is to process high-level user instructions and achieve the specified tasks with robots' primitive actions. To address this, we propose novel algorithms by utilizing semantic roles of common verbs defined in semantic dictionaries and integrating multiple open knowledge to generate task plans. Specifically, we present a new method for matching and recovering semantics of user instructions and a novel task planner that exploits functional knowledge of robot's action model. To verify and evaluate our approach, we implemented a prototype system using knowledge from several open resources. Experiments on our system confirmed the correctness and efficiency of our algorithms. Notably, our system has been deployed in the KeJia robot, which participated the annual RoboCup@Home competitions in the past three years and achieved encouragingly high scores in the benchmark tests.

cs.RO↗

arXiv2016

Deeply Exploit Depth Information for Object Detection

This paper addresses the issue on how to more effectively coordinate the depth with RGB aiming at boosting the performance of RGB-D object detection. Particularly, we investigate two primary ideas under the CNN model: property derivation and property fusion. Firstly, we propose that the depth can be utilized not only as a type of extra information besides RGB but also to derive more visual properties for comprehensively describing the objects of interest. So a two-stage learning framework consisting of property derivation and fusion is constructed. Here the properties can be derived either from the provided color/depth or their pairs (e.g. the geometry contour adopted in this paper). Secondly, we explore the fusion method of different properties in feature learning, which is boiled down to, under the CNN model, from which layer the properties should be fused together. The analysis shows that different semantic properties should be learned separately and combined before passing into the final classifier. Actually, such a detection way is in accordance with the mechanism of the primary neural cortex (V1) in brain. We experimentally evaluate the proposed method on the challenging dataset, and have achieved state-of-the-art performance.

cs.CV↗

arXiv2015

Learning High-level Prior with Convolutional Neural Networks for Semantic Segmentation

This paper proposes a convolutional neural network that can fuse high-level prior for semantic image segmentation. Motivated by humans' vision recognition system, our key design is a three-layer generative structure consisting of high-level coding, middle-level segmentation and low-level image to introduce global prior for semantic segmentation. Based on this structure, we proposed a generative model called conditional variational auto-encoder (CVAE) that can build up the links behind these three layers. These important links include an image encoder that extracts high level info from image, a segmentation encoder that extracts high level info from segmentation, and a hybrid decoder that outputs semantic segmentation from the high level prior and input image. We theoretically derive the semantic segmentation as an optimization problem parameterized by these links. Finally, the optimization problem enables us to take advantage of state-of-the-art fully convolutional network structure for the implementation of the above encoders and decoder. Experimental results on several representative datasets demonstrate our supreme performance for semantic segmentation.

cs.CV↗

arXiv2015

Non-local correction to the energy-momentum tensor for $ϕ^{3}$ theory in six dimensions

Applying the background field method, we construct by explicit computation the leading-order non-local quantum correction to the on-shell effective action for $ϕ^3$ theory in six dimensions. We then use the resulting action to obtain the non-local correction to the energy-momentum tensor. At leading order, we find that this non-local correction modifies the virial current when the scalar field is minimally coupled to gravity. This is to be compared to the classically Weyl invariant case, where it only corrects the traceless part of the energy-momentum tensor.

hep-th↗

arXiv2014

Electrodynamics Modified by Some Dimension-five Lorentz Violating Interactions: Radiative Corrections

We study radiative corrections to massless quantum electrodynamics modified by two dimension-five LV interactions $\barΨ γ^μ b'^ν F_{μν}Ψ$ and $\barΨγ^μb^ν \tilde{F}_{μν} Ψ$ in the framework of effective field theories. All divergent one-particle-irreducible Feynman diagrams are calculated at one-loop order and several related issues are discussed. It is found that massless quantum electrodynamics modified by the interaction $\barΨ γ^μ b'^ν F_{μν}Ψ$ alone is one-loop renormalizable and the result can be understood on the grounds of symmetry. In this context the one-loop Lorentz-violating beta function is derived and the corresponding running coefficients are obtained.

hep-th↗