arXiv subjects

Chun Chen

Publications and source records attributed to Chun Chen.

At least 91 records · Page 5Linked to original sources

arXiv2022

High-contrast, speckle-free, true 3D holography via binary CGH optimization

Holography is a promising approach to implement the three-dimensional (3D) projection beyond the present two-dimensional technology. True 3D holography requires abilities of arbitrary 3D volume projection with high-axial resolution and independent control of all 3D voxels. However, it has been challenging to implement the true 3D holography with high-reconstruction quality due to the speckle. Here, we propose the practical solution to realize speckle-free, high-contrast, true 3D holography by combining random-phase, temporal multiplexing, binary holography, and binary optimization. We adopt the random phase for the true 3D implementation to achieve the maximum axial resolution with fully independent control of the 3D voxels. We develop the high-performance binary hologram optimization framework to minimize the binary quantization noise, which provides accurate and high-contrast reconstructions for 2D as well as 3D cases. Utilizing the fast operation of binary modulation, the full-color high-framerate holographic video projection is realized while the speckle noise of random phase is overcome by temporal multiplexing. Our high-quality true 3D holography is experimentally verified by projecting multiple arbitrary dense images simultaneously. The proposed method can be adopted in various applications of holography, where we show additional demonstration that realistic true 3D hologram in VR and AR near-eye displays. The realization will open a new path towards the next generation of holography.

eess.IV↗

arXiv2021

Learning Dynamic Preference Structure Embedding From Temporal Networks

The dynamics of temporal networks lie in the continuous interactions between nodes, which exhibit the dynamic node preferences with time elapsing. The challenges of mining temporal networks are thus two-fold: the dynamic structure of networks and the dynamic node preferences. In this paper, we investigate the dynamic graph sampling problem, aiming to capture the preference structure of nodes dynamically in cooperation with GNNs. Our proposed Dynamic Preference Structure (DPS) framework consists of two stages: structure sampling and graph fusion. In the first stage, two parameterized samplers are designed to learn the preference structure adaptively with network reconstruction tasks. In the second stage, an additional attention layer is designed to fuse two sampled temporal subgraphs of a node, generating temporal node embeddings for downstream tasks. Experimental results on many real-life temporal networks show that our DPS outperforms several state-of-the-art methods substantially owing to learning an adaptive preference structure. The code will be released soon at https://github.com/doujiang-zheng/Dynamic-Preference-Structure.

cs.IR↗

arXiv2021

Real-time intelligent big data processing: technology, platform, and applications

Human beings keep exploring the physical space using information means. Only recently, with the rapid development of information technologies and the increasing accumulation of data, human beings can learn more about the unknown world with data-driven methods. Given data timeliness, there is a growing awareness of the importance of real-time data. There are two categories of technologies accounting for data processing: batching big data and streaming processing, which have not been integrated well. Thus, we propose an innovative incremental processing technology named after Stream Cube to process both big data and stream data. Also, we implement a real-time intelligent data processing system, which is based on real-time acquisition, real-time processing, real-time analysis, and real-time decision-making. The real-time intelligent data processing technology system is equipped with a batching big data platform, data analysis tools, and machine learning models. Based on our applications and analysis, the real-time intelligent data processing system is a crucial solution to the problems of the national society and economy.

cs.DC↗

arXiv2021

Cross-Layer Distillation with Semantic Calibration

Knowledge distillation is a technique to enhance the generalization ability of a student model by exploiting outputs from a teacher model. Recently, feature-map based variants explore knowledge transfer between manually assigned teacher-student pairs in intermediate layers for further improvement. However, layer semantics may vary in different neural networks and semantic mismatch in manual layer associations will lead to performance degeneration due to negative regularization. To address this issue, we propose Semantic Calibration for cross-layer Knowledge Distillation (SemCKD), which automatically assigns proper target layers of the teacher model for each student layer with an attention mechanism. With a learned attention distribution, each student layer distills knowledge contained in multiple teacher layers rather than a specific intermediate layer for appropriate cross-layer supervision. We further provide theoretical analysis of the association weights and conduct extensive experiments to demonstrate the effectiveness of our approach. Code is avaliable at \url{https://github.com/DefangChen/SemCKD}.

cs.CV↗

arXiv2021

SamWalker++: recommendation with informative sampling strategy

Recommendation from implicit feedback is a highly challenging task due to the lack of reliable negative feedback data. Existing methods address this challenge by treating all the un-observed data as negative (dislike) but downweight the confidence of these data. However, this treatment causes two problems: (1) Confidence weights of the unobserved data are usually assigned manually, which lack flexibility and may create empirical bias on evaluating user's preference. (2) To handle massive volume of the unobserved feedback data, most of the existing methods rely on stochastic inference and data sampling strategies. However, since a user is only aware of a very small fraction of items in a large dataset, it is difficult for existing samplers to select informative training instances in which the user really dislikes the item rather than does not know it. To address the above two problems, we propose two novel recommendation methods SamWalker and SamWalker++ that support both adaptive confidence assignment and efficient model learning. SamWalker models data confidence with a social network-aware function, which can adaptively specify different weights to different data according to users' social contexts. However, the social network information may not be available in many recommender systems, which hinders application of SamWalker. Thus, we further propose SamWalker++, which does not require any side information and models data confidence with a constructed pseudo-social network. We also develop fast random-walk-based sampling strategies for our SamWalker and SamWalker++ to adaptively draw informative training instances, which can speed up gradient estimation and reduce sampling variance. Extensive experiments on five real-world datasets demonstrate the superiority of the proposed SamWalker and SamWalker++.

cs.IR↗

arXiv2020

CoSam: An Efficient Collaborative Adaptive Sampler for Recommendation

Sampling strategies have been widely applied in many recommendation systems to accelerate model learning from implicit feedback data. A typical strategy is to draw negative instances with uniform distribution, which however will severely affect model's convergency, stability, and even recommendation accuracy. A promising solution for this problem is to over-sample the ``difficult'' (a.k.a informative) instances that contribute more on training. But this will increase the risk of biasing the model and leading to non-optimal results. Moreover, existing samplers are either heuristic, which require domain knowledge and often fail to capture real ``difficult'' instances; or rely on a sampler model that suffers from low efficiency. To deal with these problems, we propose an efficient and effective collaborative sampling method CoSam, which consists of: (1) a collaborative sampler model that explicitly leverages user-item interaction information in sampling probability and exhibits good properties of normalization, adaption, interaction information awareness, and sampling efficiency; and (2) an integrated sampler-recommender framework, leveraging the sampler model in prediction to offset the bias caused by uneven sampling. Correspondingly, we derive a fast reinforced training algorithm of our framework to boost the sampler performance and sampler-recommender collaboration. Extensive experiments on four real-world datasets demonstrate the superiority of the proposed collaborative sampler model and integrated sampler-recommender framework.

cs.IR↗

arXiv2020

Fast Adaptively Weighted Matrix Factorization for Recommendation with Implicit Feedback

Recommendation from implicit feedback is a highly challenging task due to the lack of the reliable observed negative data. A popular and effective approach for implicit recommendation is to treat unobserved data as negative but downweight their confidence. Naturally, how to assign confidence weights and how to handle the large number of the unobserved data are two key problems for implicit recommendation models. However, existing methods either pursuit fast learning by manually assigning simple confidence weights, which lacks flexibility and may create empirical bias in evaluating user's preference; or adaptively infer personalized confidence weights but suffer from low efficiency. To achieve both adaptive weights assignment and efficient model learning, we propose a fast adaptively weighted matrix factorization (FAWMF) based on variational auto-encoder. The personalized data confidence weights are adaptively assigned with a parameterized neural network (function) and the network can be inferred from the data. Further, to support fast and stable learning of FAWMF, a new specific batch-based learning algorithm fBGD has been developed, which trains on all feedback data but its complexity is linear to the number of observed data. Extensive experiments on real-world datasets demonstrate the superiority of the proposed FAWMF and its learning algorithm fBGD.

cs.IR↗

arXiv2020

Majorana corner flat bands in two-dimensional second-order topological superconductors

In this paper we find that confining a second-order topological superconductor with a harmonic potential leads to a proliferation of Majorana corner modes. As a consequence, this results in the formation of Majorana corner flat bands which have a fundamentally different origin from that of the conventional mechanism. This is due to the fact that they arise solely from the one-dimensional gapped boundary states of the hybrid system that become gapless without the bulk gap closing under the increase of the trapping potential magnitude. The Majorana corner states are found to be robust against the strength of the harmonic trap and the transition from Majorana corner states to Majorana flat bands is merely a smooth crossover. As a harmonic trap can potentially be realized in heterostructures, this proposal paves a way to observe these Majorana corner flat bands in an experimental context.

cond-mat.supr-con↗

arXiv2019

Online Knowledge Distillation with Diverse Peers

Distillation is an effective knowledge-transfer technique that uses predicted distributions of a powerful teacher model as soft targets to train a less-parameterized student model. A pre-trained high capacity teacher, however, is not always available. Recently proposed online variants use the aggregated intermediate predictions of multiple student models as targets to train each student model. Although group-derived targets give a good recipe for teacher-free distillation, group members are homogenized quickly with simple aggregation functions, leading to early saturated solutions. In this work, we propose Online Knowledge Distillation with Diverse peers (OKDDip), which performs two-level distillation during training with multiple auxiliary peers and one group leader. In the first-level distillation, each auxiliary peer holds an individual set of aggregation weights generated with an attention-based mechanism to derive its own targets from predictions of other auxiliary peers. Learning from distinct target distributions helps to boost peer diversity for effectiveness of group-based distillation. The second-level distillation is performed to transfer the knowledge in the ensemble of auxiliary peers further to the group leader, i.e., the model used for inference. Experimental results show that the proposed framework consistently gives better performance than state-of-the-art approaches without sacrificing training or inference complexity, demonstrating the effectiveness of the proposed two-level distillation framework.

cs.LG↗

arXiv2019

From supersymmetric sine-Gordon equation to the superconformal minimal model

We propose a generalization of the Grover-Sheng-Vishwanath model which, as solved by the density-matrix renormalization group, realizes the emergent supersymmetric criticality in the universality class of the even series of the $\mathcal{N}\!=\!1$ superconformal minimal models characterized by a central charge $5/4$. This chain model describes the topological phase transition of the propagating Majorana edge mode in topological superconductors coupled with the two-flavour Ising magnetic fluctuations (or the $XZ$ type). Using bosonization and perturbative renormalization group, we show that the augmented degrees of freedom trigger a paradigm shift from the supersymmetric Landau-Ginzburg action to the variant of the supersymmetric sine-Gordon equation, which, in the massless case, can flow towards the supersymmetric minimal series upon a generalized Feigin-Fuchs construction. Therefore, the present lattice model comprises a concrete system that exhibits the spacetime supersymmetry through the distinct route.

cond-mat.str-el↗

arXiv2019

Revisiting the Ramond sector of the $\mathcal{N}\!=\!1$ superconformal minimal models

Key to the exact solubility of the unitary minimal models in two-dimensional conformal field theory is the organization of their Hilbert space into Verma modules, whereby all eigenstates of the Hamiltonian are obtained by the repeated action of Virasoro lowering operators onto a finite set of highest-weight states. The usual representation-theoretic approach to removing from all modules zero-norm descendant states generated in such a way is based on the assumption that those states form a nested sequence of Verma submodules built upon singular vectors, i.e., descendant highest-weight states. We show that this fundamental assumption breaks down for the Ramond-sector Verma module with highest weight $c/24$ in the even series of $\mathcal{N}\!=\!1$ superconformal minimal models with central charge $c$. To resolve this impasse, we conjecture, and prove at low orders, the existence of a nested sequence of linear-dependence relations that enables us to compute the character of the irreducible $c/24$ module. Based on this character formula, we argue that imposing modular invariance of the torus partition function requires the introduction of a non-null odd-parity Ramond-sector ground state. This symmetrization of the ground-state manifold allows us to uncover a set of conformally invariant boundary conditions not previously discussed and absent in the odd series of superconformal minimal models, and to derive for the first time a complete set of fusion rules for the even series of those models.

hep-th↗

arXiv2017

How does a locally constrained quantum system localize?

At low energy, the dynamics of excitations of many physical systems are locally constrained. Examples include frustrated anti-ferromagnets, fractional quantum Hall fluids and Rydberg atoms in the blockaded regime. Can such locally constrained systems be fully many-body localized (MBL)? In this article, we answer this question affirmatively and elucidate the structure of the accompanying quasi-local integrals of motion. By studying disordered spin chains subject to a projection constraint in the $z$-direction, we show that full MBL is stable at strong $z$-field disorder and identify a new mechanism of localization through resonance at strong transverse disorder. However MBL is not guaranteed; the constraints can `frustrate' the tendency of the spins to align with the transverse fields and lead to full thermalization or criticality. We further provide evidence that the transition is discontinuous in local observables with large sample-to-sample variations. Our study has direct consequences for current quench experiments in Rydberg atomic chains.

cond-mat.dis-nn↗

arXiv2017

Flux-Stabilized Majorana Zero Modes in Coupled One-Dimensional Fermi Wires

One promising avenue to study one-dimensional ($1$D) topological phases is to realize them in synthetic materials such as cold atomic gases. Intriguingly, it is possible to realize Majorana boundary modes in a $1$D number-conserving system consisting of two fermionic chains coupled only by pair-hopping processes. It is commonly believed that significant interchain single-particle tunneling necessarily destroys these Majorana modes, as it spoils the $\mathbb{Z}_2$ fermion parity symmetry that protects them. In this Letter, we present a new mechanism to overcome this obstacle, by piercing a (synthetic) magnetic $π$-flux through each plaquette of the Fermi ladder. Using bosonization, we show that in this case there exists an exact leg-interchange symmetry that is robust to interchain hopping, and acts as fermion parity at long wavelengths. We utilize density matrix renormalization group and exact diagonalization to verify that the resulting model exhibits Majorana boundary modes up to large single-particle tunnelings, comparable to the intrachain hopping strength. Our work highlights the unusual impacts of different topologically trivial band structures on these interaction-driven topological phases, and identifies a distinct route to stabilizing Majorana boundary modes in $1$D fermionic ladders.

cond-mat.str-el↗

arXiv2016

Relational Multi-Manifold Co-Clustering

Co-clustering targets on grouping the samples (e.g., documents, users) and the features (e.g., words, ratings) simultaneously. It employs the dual relation and the bilateral information between the samples and features. In many realworld applications, data usually reside on a submanifold of the ambient Euclidean space, but it is nontrivial to estimate the intrinsic manifold of the data space in a principled way. In this study, we focus on improving the co-clustering performance via manifold ensemble learning, which is able to maximally approximate the intrinsic manifolds of both the sample and feature spaces. To achieve this, we develop a novel co-clustering algorithm called Relational Multi-manifold Co-clustering (RMC) based on symmetric nonnegative matrix tri-factorization, which decomposes the relational data matrix into three submatrices. This method considers the intertype relationship revealed by the relational data matrix, and also the intra-type information reflected by the affinity matrices encoded on the sample and feature data distributions. Specifically, we assume the intrinsic manifold of the sample or feature space lies in a convex hull of some pre-defined candidate manifolds. We want to learn a convex combination of them to maximally approach the desired intrinsic manifold. To optimize the objective function, the multiplicative rules are utilized to update the submatrices alternatively. Besides, both the entropic mirror descent algorithm and the coordinate descent algorithm are exploited to learn the manifold coefficient vector. Extensive experiments on documents, images and gene expression data sets have demonstrated the superiority of the proposed algorithm compared to other well-established methods.

cs.LG↗

arXiv2016

Spectral Graph Cut from a Filtering Point of View

Spectral graph theory is well known and widely used in computer vision. In this paper, we analyze image segmentation algorithms that are based on spectral graph theory, e.g., normalized cut, and show that there is a natural connection between spectural graph theory based image segmentationand and edge preserving filtering. Based on this connection we show that the normalized cut algorithm is equivalent to repeated iterations of bilateral filtering. Then, using this equivalence we present and implement a fast normalized cut algorithm for image segmentation. Experiments show that our implementation can solve the original optimization problem in the normalized cut algorithm 10 to 100 times faster. Furthermore, we present a new algorithm called conditioned normalized cut for image segmentation that can easily incorporate color image patches and demonstrate how this segmentation problem can be solved with edge preserving filtering.

cs.CV↗

arXiv2016

Inhomogeneous Topological Superfluidity in One-Dimensional Spin-Orbit-Coupled Fermi Gases

We theoretically predict an exotic topological superfluid state with spatially modulated pairing gap in one-dimensional spin-orbit-coupled Fermi gases. This inhomogeneous topological superfluidity is induced by applying simultaneously a perpendicular Zeeman magnetic field and an equally weighted Rashba and Dresselhaus spin-orbit coupling in one-dimensional optical lattices. Based on the self-consistent Bogoliubov--de Gennes theory, we confirm that this novel topological phase is a unique condensation of Cooper pairs, which manifests the interplay between the inhomogeneity of superfluid and its nontrivial topological structure. The properties of the emergent Majorana bound states are investigated in detail by examining the associated $\mathbb{Z}_{2}$ topological number, the eigenenergy and density of states spectra, as well as the wave functions of the localized Majorana end modes. Experimental feasibility of observing this new topological state of matter is also discussed.

cond-mat.quant-gas↗

arXiv2016

Scalable Image Retrieval by Sparse Product Quantization

Fast Approximate Nearest Neighbor (ANN) search technique for high-dimensional feature indexing and retrieval is the crux of large-scale image retrieval. A recent promising technique is Product Quantization, which attempts to index high-dimensional image features by decomposing the feature space into a Cartesian product of low dimensional subspaces and quantizing each of them separately. Despite the promising results reported, their quantization approach follows the typical hard assignment of traditional quantization methods, which may result in large quantization errors and thus inferior search performance. Unlike the existing approaches, in this paper, we propose a novel approach called Sparse Product Quantization (SPQ) to encoding the high-dimensional feature vectors into sparse representation. We optimize the sparse representations of the feature vectors by minimizing their quantization errors, making the resulting representation is essentially close to the original data in practice. Experiments show that the proposed SPQ technique is not only able to compress data, but also an effective encoding technique. We obtain state-of-the-art results for ANN search on four public image datasets and the promising results of content-based image retrieval further validate the efficacy of our proposed method.

cs.CV↗

arXiv2016

Type-II Topological Meissner States

We study the \emph{orbital effects} of the synthetic magnetic fields in an interacting square lattice two-leg fermionic ladder model with the number-conserving pair hopping that hosting the Majorana bound states. By utilizing density matrix renormalization group and exact diagonalization, we identify a novel type-II topological Meissner $($\emph{topo}-Meissner$)$ phase $($as distinguished from the low-field type-I \emph{topo}-Meissner state$)$ when threading a \emph{high} magnetic flux through the plaquette of the ladder, which not only exhibits a large uniformly circulating chiral current along the legs, the characteristics of the celebrated Meissner state, but also accommodates a topologically protected ground-state manifold due to the \emph{reentrant} emergence of the Majorana end modes. Our work reveals some interesting interference effects resulting from the interplay between the gauge fields and the strong interactions in establishing the intrinsic topological states of matter in low-dimensional quantum systems.

cond-mat.str-el↗