Search

Fine-grained sketch-based image retrieval (FG-SBIR) addresses the problem of retrieving a particular photo instance given a user's query sketch. Its widespread applicability is however hindered by the fact that drawing a sketch takes time, and most people struggle to draw a complete and faithful sketch. In this paper, we reformulate the conventional FG-SBIR framework to tackle these challenges, with the ultimate goal of retrieving the target photo with the least number of strokes possible. We further propose an on-the-fly design that starts retrieving as soon as the user starts drawing. To accomplish this, we devise a reinforcement learning-based cross-modal retrieval framework that directly optimizes rank of the ground-truth photo over a complete sketch drawing episode. Additionally, we introduce a novel reward scheme that circumvents the problems related to irrelevant sketch strokes, and thus provides us with a more consistent rank list during the retrieval. We achieve superior early-retrieval efficiency over state-of-the-art methods and alternative baselines on two publicly available fine-grained sketch retrieval datasets., Comment: IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2020 [Oral Presentation] Code: https://github.com/AyanKumarBhunia/on-the-fly-FGSBIR

Existing sketch-analysis work studies sketches depicting static objects or scenes. In this work, we propose a novel cross-modal retrieval problem of fine-grained instance-level sketch-based video retrieval (FG-SBVR), where a sketch sequence is used as a query to retrieve a specific target video instance. Compared with sketch-based still image retrieval, and coarse-grained category-level video retrieval, this is more challenging as both visual appearance and motion need to be simultaneously matched at a fine-grained level. We contribute the first FG-SBVR dataset with rich annotations. We then introduce a novel multi-stream multi-modality deep network to perform FG-SBVR under both strong and weakly supervised settings. The key component of the network is a relation module, designed to prevent model over-fitting given scarce training data. We show that this model significantly outperforms a number of existing state-of-the-art models designed for video analysis.

Decentralized learning has gained great popularity to improve learning efficiency and preserve data privacy. Each computing node makes equal contribution to collaboratively learn a Deep Learning model. The elimination of centralized Parameter Servers (PS) can effectively address many issues such as privacy, performance bottleneck and single-point-failure. However, how to achieve Byzantine Fault Tolerance in decentralized learning systems is rarely explored, although this problem has been extensively studied in centralized systems. In this paper, we present an in-depth study towards the Byzantine resilience of decentralized learning systems with two contributions. First, from the adversarial perspective, we theoretically illustrate that Byzantine attacks are more dangerous and feasible in decentralized learning systems: even one malicious participant can arbitrarily alter the models of other participants by sending carefully crafted updates to its neighbors. Second, from the defense perspective, we propose UBAR, a novel algorithm to enhance decentralized learning with Byzantine Fault Tolerance. Specifically, UBAR provides a Uniform Byzantine-resilient Aggregation Rule for benign nodes to select the useful parameter updates and filter out the malicious ones in each training iteration. It guarantees that each benign node in a decentralized system can train a correct model under very strong Byzantine attacks with an arbitrary number of faulty nodes. We conduct extensive experiments on standard image classification tasks and the results indicate that UBAR can effectively defeat both simple and sophisticated Byzantine attacks with higher performance efficiency than existing solutions.

Existing meta-learning based few-shot learning (FSL) methods typically adopt an episodic training strategy whereby each episode contains a meta-task. Across episodes, these tasks are sampled randomly and their relationships are ignored. In this paper, we argue that the inter-meta-task relationships should be exploited and those tasks are sampled strategically to assist in meta-learning. Specifically, we consider the relationships defined over two types of meta-task pairs and propose different strategies to exploit them. (1) Two meta-tasks with disjoint sets of classes: this pair is interesting because it is reminiscent of the relationship between the source seen classes and target unseen classes, featured with domain gap caused by class differences. A novel learning objective termed meta-domain adaptation (MDA) is proposed to make the meta-learned model more robust to the domain gap. (2) Two meta-tasks with identical sets of classes: this pair is useful because it can be employed to learn models that are robust against poorly sampled few-shots. To that end, a novel meta-knowledge distillation (MKD) objective is formulated. There are some mistakes in the experiments. We thus choose to withdraw this paper., Comment: There are some mistakes in the experiments. We thus choose to withdraw this paper

To recognize the unseen classes with only few samples, few-shot learning (FSL) uses prior knowledge learned from the seen classes. A major challenge for FSL is that the distribution of the unseen classes is different from that of those seen, resulting in poor generalization even when a model is meta-trained on the seen classes. This class-difference-caused distribution shift can be considered as a special case of domain shift. In this paper, for the first time, we propose a domain adaptation prototypical network with attention (DAPNA) to explicitly tackle such a domain shift problem in a meta-learning framework. Specifically, armed with a set transformer based attention module, we construct each episode with two sub-episodes without class overlap on the seen classes to simulate the domain shift between the seen and unseen classes. To align the feature distributions of the two sub-episodes with limited training samples, a feature transfer network is employed together with a margin disparity discrepancy (MDD) loss. Importantly, theoretical analysis is provided to give the learning bound of our DAPNA. Extensive experiments show that our DAPNA outperforms the state-of-the-art FSL alternatives, often by significant margins., Comment: There exist some mistakes in the experiments

Person re-identification (Re-ID) aims at retrieving a person of interest across multiple non-overlapping cameras. With the advancement of deep neural networks and increasing demand of intelligent video surveillance, it has gained significantly increased interest in the computer vision community. By dissecting the involved components in developing a person Re-ID system, we categorize it into the closed-world and open-world settings. The widely studied closed-world setting is usually applied under various research-oriented assumptions, and has achieved inspiring success using deep learning techniques on a number of datasets. We first conduct a comprehensive overview with in-depth analysis for closed-world person Re-ID from three different perspectives, including deep feature representation learning, deep metric learning and ranking optimization. With the performance saturation under closed-world setting, the research focus for person Re-ID has recently shifted to the open-world setting, facing more challenging issues. This setting is closer to practical applications under specific scenarios. We summarize the open-world Re-ID in terms of five different aspects. By analyzing the advantages of existing methods, we design a powerful AGW baseline, achieving state-of-the-art or at least comparable performance on twelve datasets for FOUR different Re-ID tasks. Meanwhile, we introduce a new evaluation metric (mINP) for person Re-ID, indicating the cost for finding all the correct matches, which provides an additional criteria to evaluate the Re-ID system for real applications. Finally, some important yet under-investigated open issues are discussed., Comment: 20 pages, 8 figures. Accepted by IEEE TPAMI

Free-hand sketches are highly illustrative, and have been widely used by humans to depict objects or stories from ancient times to the present. The recent prevalence of touchscreen devices has made sketch creation a much easier task than ever and consequently made sketch-oriented applications increasingly popular. The progress of deep learning has immensely benefited free-hand sketch research and applications. This paper presents a comprehensive survey of the deep learning techniques oriented at free-hand sketch data, and the applications that they enable. The main contents of this survey include: (i) A discussion of the intrinsic traits and unique challenges of free-hand sketch, to highlight the essential differences between sketch data and other data modalities, e.g., natural photos. (ii) A review of the developments of free-hand sketch research in the deep learning era, by surveying existing datasets, research topics, and the state-of-the-art methods through a detailed taxonomy and experimental evaluation. (iii) Promotion of future work via a discussion of bottlenecks, open problems, and potential research directions for the community., Comment: This paper is accepted by IEEE TPAMI

TThe transition temperature (TC) between normal and superconducting states usually exhibits a dramatic increase or decrease with increasing applied pressure. Here we present, in contrast, a new kind of superconductor that exhibits the exotic feature that TC is robust against large volume shrinkages induced by applied pressure (here naming them as "RSAVS superconductors"). Extraordinarily, the TC in these materials stays almost constant over a large pressure range, e.g. over 136 GPa in the (TaNb)0.67(HfZrTi)0.33 high entropy alloy and 141 GPa in the NbTi commercial alloy. We show that the RSAVS behavior also exists in another high entropy alloy (ScZrNbTa)0.6(RhPd)0.4, and in superconducting elemental Ta and Nb, indicating that this behavior, which has never previously been identified or predicted by theory, occurs universally in some conventional superconductors. Our electronic structure calculations indicate that although the electronic density of state (DOS) at the Fermi level in the RSAVS state is dominated by the electrons from the degenerate dxy, dxz and dyz orbitals, these electrons decrease in influence with increasing pressure. In contrast, however, the contribution of the degenerate dx2-y2 and dz2 orbital electrons remains almost unchanged at the Fermi level, suggesting that these are the electrons that may play a crucial role in stabilizing the TC in the RSAVS state., Comment: 12 pages, 4 figures

Two-dimensional materials offer opportunities for unravelling unprecedented ordered states at single layer limit. Among such ordered states, Mott phase is rarely explored. Here, we report the Mott phase in van der Waals chromium (II) iodide (CrI2) films. High quality CrI2 films with atomically flat surface and macro size are grown on graphitized 6H-SiC(0001) substrate by molecular beam epitaxy. By in situ low temperature scanning tunneling microscopy and spectroscopy (STM/STS), we reveal that the film has a band gap as large as ~3.2 eV, which is nearly thickness independent. Density functional plus dynamic mean field theory calculations suggest that CrI2 films may be a strong Mott insulator with a ferromagnetically ordered ground state. The Mott phase is corroborated by the spectral band splitting, that is consistent with the extended Hubbard model, and gap reduction at charge dopants. Our study provides a platform for studying correlated electron states at single layer limit., Comment: 18 pages, 4 figures

Tensor renormalization group (TRG) constitutes an important methodology for accurate simulations of strongly correlated lattice models. Facilitated by the automatic differentiation technique widely used in deep learning, we propose a uniform framework of differentiable TRG ($\partial$TRG) that can be applied to improve various TRG methods, in an automatic fashion. Essentially, $\partial$TRG systematically extends the concept of second renormalization [PRL 103, 160601 (2009)] where the tensor environment is computed recursively in the backward iteration, in the sense that given the forward process of TRG, $\partial$TRG automatically finds the gradient through backpropagation, with which one can deeply "train" the tensor networks. We benchmark $\partial$TRG in solving the square-lattice Ising model, and demonstrate its power by simulating one- and two-dimensional quantum systems at finite temperature. The deep optimization as well as GPU acceleration renders $\partial$TRG manybody simulations with high efficiency and accuracy., Comment: 13 pages, 9 figures

Coupling between $\sigma$-bonding electrons and phonons is generally very strong. To metallize $\sigma$-electrons provides a promising route to hunt for new high-T$_c$ superconductors. Based on this picture and first-principles density functional calculation with Wannier interpolation for electronic structure and lattice dynamics, we predict that trilayer film LiB$_2$C$_2$ is a good candidate to realize this kind of high-T$_c$ superconductivity. By solving the anisotropic Eliashberg equations, we find that free-standing trilayer LiB$_2$C$_2$ is a phonon-mediated superconductor with T$_c$ exceeding the liquid-nitrogen temperature at ambient pressure. The transition temperature can be further raised to 125 K by applying a biaxial tensile strain., Comment: 5 pages, 6 figures

The iron-based superconductor FeSe has attracted much recent attention because of its simple crystal structure, distinct electronic structure and rich physics exhibited by itself and its derivatives. Determination of its intrinsic electronic structure is crucial to understand its physical properties and superconductivity mechanism. Both theoretical and experimental studies so far have provided a picture that FeSe consists of one hole-like Fermi surface around the Brillouin zone center in its nematic state. Here we report direct observation of two hole-like Fermi surface sheets around the Brillouin zone center, and the splitting of the associated bands, in the nematic state of FeSe by taking high resolution laser-based angle-resolved photoemission measurements. These results indicate that, in addition to nematic order and spin-orbit coupling, there is an additional order in FeSe that breaks either inversion or time reversal symmetries. The new Fermi surface topology asks for reexamination of the existing theoretical and experimental understanding of FeSe and stimulates further efforts to identify the origin of the hidden order in its nematic state.

High-resolution laser-based angle-resolved photoemission measurements have been carried out on Bi$_2$Sr$_2$CaCu$_2$O$_{8+\delta}$ (Bi2212) and Bi$_2$Sr$_{2-x}$La$_x$CuO$_{6+\delta}$ (Bi2201) superconductors. Unexpected hybridization between the main band and the superstructure band in Bi2212 is clearly revealed. In the momentum space where one main Fermi surface intersects with one superstructure Fermi surface, four bands are observed instead of two. The hybridization exists in both superconducting state and normal state, and in Bi2212 samples with different doping levels. Such a hybridization is not observed in Bi2201. This phenomenon can be understood by considering the bilayer splitting in Bi2212, the selective hybridization of two bands with peculiar combinations, and the altered matrix element effects of the hybridized bands. These observations provide strong evidence on the origin of the superstructure band which is intrinsic to the CuO$_2$ planes. Therefore, understanding physical properties and superconductivity mechanism in Bi2212 should consider the complete Fermi surface topology which involves the main bands, the superstructure bands and their interactions., Comment: 20 pages, 4 figures

Person re-identification (re-ID), which aims to re-identify people across different camera views, has been significantly advanced by deep learning in recent years, particularly with convolutional neural networks (CNNs). In this paper, we present Torchreid, a software library built on PyTorch that allows fast development and end-to-end training and evaluation of deep re-ID models. As a general-purpose framework for person re-ID research, Torchreid provides (1) unified data loaders that support 15 commonly used re-ID benchmark datasets covering both image and video domains, (2) streamlined pipelines for quick development and benchmarking of deep re-ID models, and (3) implementations of the latest re-ID CNN architectures along with their pre-trained models to facilitate reproducibility as well as future research. With a high-level modularity in its design, Torchreid offers a great flexibility to allow easy extension to new datasets, CNN models and loss functions., Comment: Tech report

An effective person re-identification (re-ID) model should learn feature representations that are both discriminative, for distinguishing similar-looking people, and generalisable, for deployment across datasets without any adaptation. In this paper, we develop novel CNN architectures to address both challenges. First, we present a re-ID CNN termed omni-scale network (OSNet) to learn features that not only capture different spatial scales but also encapsulate a synergistic combination of multiple scales, namely omni-scale features. The basic building block consists of multiple convolutional streams, each detecting features at a certain scale. For omni-scale feature learning, a unified aggregation gate is introduced to dynamically fuse multi-scale features with channel-wise weights. OSNet is lightweight as its building blocks comprise factorised convolutions. Second, to improve generalisable feature learning, we introduce instance normalisation (IN) layers into OSNet to cope with cross-dataset discrepancies. Further, to determine the optimal placements of these IN layers in the architecture, we formulate an efficient differentiable architecture search algorithm. Extensive experiments show that, in the conventional same-dataset setting, OSNet achieves state-of-the-art performance, despite being much smaller than existing re-ID models. In the more challenging yet practical cross-dataset setting, OSNet beats most recent unsupervised domain adaptation methods without using any target data. Our code and models are released at \texttt{https://github.com/KaiyangZhou/deep-person-reid}., Comment: TPAMI 2021. Journal extension of arXiv:1905.00953. Updates: added appendix. arXiv admin note: text overlap with arXiv:1905.00953

It has long been a challenging task to find compounds with similar crystal and electronic structures as cuprate superconductors with low dimensionality and strong antiferromagnetic fluctuations. The parent compounds of cuprate superconductors are Mott insulators with strong in-plane antiferromagnetic exchange interactions between Cu moments. Here we show, based on first-principles density functional calculations, that AFeSe2 (A=Tl, K, Rb, or Cs) exhibit many of the physical properties common to the cuprate parent compounds: (1) the FeSe2 layer in AFeSe2 is similar in crystalline and electronic structures to the CuO2 plane in cuprates, although Se atoms are not coplanar to the square Fe-lattice; (2) they are antiferromagnetic insulators, but with relatively small charge excitation gaps; (3) their ground states are Neel antiferromagnetic ordered, similar as in cuprates; and (4) the antiferromagnetic exchange interactions between Fe moments are larger than in other iron-based superconducting materials, but comparable to those in cuprates. Like cuprates, these compounds may become high-Tc superconductors upon doping of charge carriers either by chemical substitution or intercalation or by liquid or solid gating., Comment: 6 pages, 4 figures, 2 tables

In this paper, we propose to tackle the challenging few-shot learning (FSL) problem by learning global class representations using both base and novel class training samples. In each training episode, an episodic class mean computed from a support set is registered with the global representation via a registration module. This produces a registered global class representation for computing the classification loss using a query set. Though following a similar episodic training pipeline as existing meta learning based approaches, our method differs significantly in that novel class training samples are involved in the training from the beginning. To compensate for the lack of novel class training samples, an effective sample synthesis strategy is developed to avoid overfitting. Importantly, by joint base-novel class training, our approach can be easily extended to a more practical yet challenging FSL setting, i.e., generalized FSL, where the label space of test data is extended to both base and novel classes. Extensive experiments show that our approach is effective for both of the two FSL settings., Comment: Accepted by ICCV2019

Automatic data abstraction is an important capability for both benchmarking machine intelligence and supporting summarization applications. In the former one asks whether a machine can `understand' enough about the meaning of input data to produce a meaningful but more compact abstraction. In the latter this capability is exploited for saving space or human time by summarizing the essence of input data. In this paper we study a general reinforcement learning based framework for learning to abstract sequential data in a goal-driven way. The ability to define different abstraction goals uniquely allows different aspects of the input data to be preserved according to the ultimate purpose of the abstraction. Our reinforcement learning objective does not require human-defined examples of ideal abstraction. Importantly our model processes the input sequence holistically without being constrained by the original input order. Our framework is also domain agnostic -- we demonstrate applications to sketch, video and text data and achieve promising results in all domains., Comment: Accepted at ICCV 2019

To decipher the mechanism of high temperature superconductivity, it is important to know how the superconducting pairing emerges from the unusual normal states of cuprate superconductors, including pseudogap, anomalous Fermi liquid and strange metal (SM). A long-standing issue under debate is how the superconducting pairing is formed and condensed in the SM phase because the superconducting transition temperature is the highest in this phase. Here, we report the first experimental observation of a pressure-induced crossover from two- to three-dimensional superconducting states in the optimally-doped Bi2Sr2CaCu2O8+delta bulk superconductor at a pressure above 2.8 GPa, through state-of-the-art in-situ high-pressure measurements of resistance, magnetoresistance and magnetic susceptibility. By analyzing the temperature dependence of resistance, we find that the two-dimensional (2D) superconducting transition exhibits a Berezinski-Kosterlitz-Thouless-like behavior. The emergence of this 2D superconducting transition provides direct and strong evidence that the SM state is predominantly 2D-like. This is important to a thorough understanding of the phase diagram of cuprate superconductors., Comment: 18 pages, 4 figures

Aiming at overcoming some inherent drawbacks and bottlenecks encountered by the conventional Knowledge, Recommendation, Search, and Social Systems, in this article we introduce the Knowledge Network System (KNS), a novel type of knowledge graph which is constructed by a new proposed algorithm, the Evolutionary Collective Intelligence (ECI) algorithm. The ECI, an agent-machine interactive algorithm, constructs the KNS by iteratively recommending interesting/matched samples/files to the agents, and meanwhile taking advantages of the collective intelligence of the agents. The ECI based KNS, to the best of our knowledge, is the first attempt in literature that integrates the functions of knowledge network construction, high-quality recommendation, new types of search and social in a same framework. Some real and potential applications of KNS and ECI are discussed, and a real system named VISVA is provided to demonstrate their efficacy. Some open problems for future works are also summarized in the end.

Manipulating the superconducting states of high-T_c cuprate superconductors in an efficient and reliable way is of great importance for their applications in next-generation electronics. Traditional methods are mostly based on a trial-and-error method that is difficult to implement and time consuming. Here, employing ionic liquid gating, a selective control of volatile and non-volatile superconductivity is achieved in pristine insulating Pr_2CuO_{4\pm\delta} film, based on two distinct mechanisms: 1) with positive electric fields, the film can be reversibly switched between non-superconducting and superconducting states, attributed to the carrier doping effect. 2) The film becomes more resistive by applying negative bias voltage up to -4 V, but strikingly, a non-volatile superconductivity is achieved once the gate voltage is removed. Such a persistent superconducting state represents a novel phenomenon in copper oxides, resulting from the doping healing of oxygen vacancies in copper-oxygen planes as unraveled by high-resolution scanning transmission electron microscope and in-situ x-ray diffraction experiments. The effective manipulation and mastering of volatile/non-volatile superconductivity in the same parent cuprate opens the door to more functionalities for superconducting electronics, as well as supplies flexible samples for investigating the nature of quantum phase transitions in high-T_c superconductors.

We have extended the canonical tree tensor network (TTN) method, which was initially introduced to simulate the zero-temperature properties of quantum lattice models on the Bethe lattice, to finite temperature simulations. By representing the thermal density matrix with a canonicalized tree tensor product operator, we optimize the TTN and accurately evaluate the thermodynamic quantities, including the internal energy, specific heat, and the spontaneous magnetization, etc, at various temperatures. By varying the anisotropic coupling constant $\Delta$, we obtain the phase diagram of the spin-1/2 Heisenberg XXZ model on the Bethe lattice, where three kinds of magnetic ordered phases, namely the ferromagnetic, XY and antiferromagnetic ordered phases, are found in low temperatures and separated from the high-$T$ paramagnetic phase by a continuous thermal phase transition at $T_c$. The XY phase is separated from the other two phases by two first-order phase transition lines at the symmetric coupling points $ \Delta=\pm 1$. We have also carried out a linear spin wave calculation on the Bethe lattice, showing that the low-energy magnetic excitations are always gapped, and find the obtained magnon gaps in very good agreement with those estimated from the TTN simulations. Despite the gapped excitation spectrum, Goldstone-like transverse fluctuation modes, as a manifestation of spontaneous continuous symmetry breaking, are observed in the ordered magnetic phases with $|\Delta|\le 1$. One remarkable feature there is that the prominent transverse correlation length reaches $\xi_c=1/\ln{(z-1)}$ for $T\leq T_c$, the maximal value allowed on a $z$-coordinated Bethe lattice., Comment: 15 pages, 13 figures

Motivated by the recent surge of field-driven phenomena discussed for Kitaev materials, in particular the experimental observation of a finite thermal Hall effect and theoretical proposals for the emergence of additional spin liquid phases beyond the conventional Kitaev spin liquid, we develop a theoretical understanding of the thermal Hall effect in honeycomb Kitaev materials in magnetic fields. Our focus is on gapless U(1) spin liquids with a spinon Fermi surface, which have been shown to arise as field-induced phases. We demonstrate that in the presence of symmetry-allowed, second-neighbor Dzyaloshinskii-Moriya interactions these spin liquids give rise to a finite, non-quantized, thermal Hall conductance in a magnetic field. The microscopic origin of this thermal Hall effect can be traced back to an interplay of Dzyaloshinskii-Moriya interaction and Zeeman coupling, which generates an internal U(1) gauge flux that twists the motion of the emergent spinons. We argue that such a non-quantized thermal Hall effect is a generic response in Kitaev models for a range of couplings., Comment: 8+4 pages, 8+1 figures

Spinel compounds have demonstrated rich functionalities but rarely shown superconductivity. Here, we report the emergence of superconductivity in the spinel $\mathrm{MgTi}_2\mathrm{O}_4$, known to be an insulator with a complicated order. The superconducting transition is achieved by engineering a superlattice of $\mathrm{MgTi}_2\mathrm{O}_4$ and $\mathrm{SrTiO}_3$. The onset transition temperature in the $\mathrm{MgTi}_2\mathrm{O}_4$ layer can be tuned from 0 to 5 K in such geometry, concurrently with a stretched $c$-axis (from 8.51 to 8.53 \AA) compared to the bulk material. Such a positive correlation without saturation suggests ample room for the further enhancement. Intriguingly, the superlattice exhibits isotropic upper critical field $H_{\mathrm{c}2}$ that breaks the Pauli limit, distinct from the highly anisotropic feature of interface superconductivity. The origin of superconductivity in the $\mathrm{MgTi}_2\mathrm{O}_4$ layer is understood in combination with the electron energy loss spectra and the first-principles electronic structure calculations, which point to the birth of superconductivity in the $\mathrm{MgTi}_2\mathrm{O}_4$ layer by preventing the Ti-Ti dimerization. Our discovery not only provides a platform to explore the interplay between the superconductivity and other exotic states, but also opens a new window to realize superconductivity in the spinel compounds as well as other titanium oxides., Comment: 5 pages, 4 figures

As an instance-level recognition problem, person re-identification (ReID) relies on discriminative features, which not only capture different spatial scales but also encapsulate an arbitrary combination of multiple scales. We call features of both homogeneous and heterogeneous scales omni-scale features. In this paper, a novel deep ReID CNN is designed, termed Omni-Scale Network (OSNet), for omni-scale feature learning. This is achieved by designing a residual block composed of multiple convolutional streams, each detecting features at a certain scale. Importantly, a novel unified aggregation gate is introduced to dynamically fuse multi-scale features with input-dependent channel-wise weights. To efficiently learn spatial-channel correlations and avoid overfitting, the building block uses pointwise and depthwise convolutions. By stacking such block layer-by-layer, our OSNet is extremely lightweight and can be trained from scratch on existing ReID benchmarks. Despite its small model size, OSNet achieves state-of-the-art performance on six person ReID datasets, outperforming most large-sized models, often by a clear margin. Code and models are available at: \url{https://github.com/KaiyangZhou/deep-person-reid}., Comment: ICCV 2019; This version adds additional training recipes for practitioners

A deep neural network is a parametrization of a multilayer mapping of signals in terms of many alternatively arranged linear and nonlinear transformations. The linear transformations, which are generally used in the fully connected as well as convolutional layers, contain most of the variational parameters that are trained and stored. Compressing a deep neural network to reduce its number of variational parameters but not its prediction power is an important but challenging problem toward the establishment of an optimized scheme in training efficiently these parameters and in lowering the risk of overfitting. Here we show that this problem can be effectively solved by representing linear transformations with matrix product operators (MPOs), which is a tensor network originally proposed in physics to characterize the short-range entanglement in one-dimensional quantum states. We have tested this approach in five typical neural networks, including FC2, LeNet-5, VGG, ResNet, and DenseNet on two widely used data sets, namely, MNIST and CIFAR-10, and found that this MPO representation indeed sets up a faithful and efficient mapping between input and output signals, which can keep or even improve the prediction accuracy with a dramatically reduced number of parameters. Our method greatly simplifies the representations in deep learning, and opens a possible route toward establishing a framework of modern neural networks which might be simpler and cheaper, but more efficient., Comment: 8+9 pages, 3+7 figures, 2+11 tables

The deconfined quantum critical point (DQCP) was originally proposed as a continuous transition between two spontaneous symmetry breaking phases in 2D spin-1/2 systems. While great efforts have been spent on the DQCP for 2D systems, both theoretically and numerically, ambiguities among the nature of the transition are still not completely clarified. Here we shift the focus to a recently proposed 1D incarnation of DQCP in a spin-1/2 chain. By solving it with the variational matrix product state in the thermodynamic limit, a continuous transition between a valence-bond solid phase and a ferromagnetic phase is discovered. The scaling dimensions of various operators are calculated and compared with those from field theoretical description. At the critical point, two emergent O(2) symmetries are revealed, and the associated conserved current operators with exact integer scaling dimensions are determined with scrutiny. Our findings provide the low-dimensional analog of DQCP where unbiased numerical results are in perfect agreement with the controlled field theoretical predictions and have extended the realm of the unconventional phase transition as well as its identification with the advanced numerical methodology., Comment: 9 pages + references + appendix, 9 figures

Differentiable programming is a fresh programming paradigm which composes parameterized algorithmic components and trains them using automatic differentiation (AD). The concept emerges from deep learning but is not only limited to training neural networks. We present theory and practice of programming tensor network algorithms in a fully differentiable way. By formulating the tensor network algorithm as a computation graph, one can compute higher order derivatives of the program accurately and efficiently using AD. We present essential techniques to differentiate through the tensor networks contractions, including stable AD for tensor decomposition and efficient backpropagation through fixed point iterations. As a demonstration, we compute the specific heat of the Ising model directly by taking the second order derivative of the free energy obtained in the tensor renormalization group calculation. Next, we perform gradient based variational optimization of infinite projected entangled pair states for quantum antiferromagnetic Heisenberg model and obtain start-of-the-art variational energy and magnetization with moderate efforts. Differentiable programming removes laborious human efforts in deriving and implementing analytical gradients for tensor network programs, which opens the door to more innovations in tensor network algorithms and applications., Comment: Typos corrected, discussion and refs added; revised version accepted for publication in PRX. Source code available at https://github.com/wangleiphy/tensorgrad

Recently, a honeycomb borophene was reported to grow successfully on Al(111) surface. Since the metallic $\sigma$-bonding bands of honeycomb boron sheet play a crucial role in the 39 K superconductivity of MgB$_2$, it is physically interesting to examine whether similar property exists in this material. We have calculated the electronic structures and the electron-phonon coupling for honeycomb borophene by explicitly considering the substrate effect using first-principles density functional theory in conjunction with the Wannier interpolation technique. We find that the $sp^2$-hybridized $\sigma$-bonding bands of honeycomb borophene are metallized due to moderate charge transfer from the Al substrate, similar as in MgB$_2$. However, the electron-phonon coupling in honeycomb borophene is much weaker than in MgB$_2$ due to the hardening of the bond-stretching boron phonon modes and the reduction of phonon density of states. Nevertheless, the interlayer coupling between Al-associated phonons and electrons in borophene is strong. Based on this observation, we predict that a 6.5 K superconducting transition can be observed in a free-standing borophene decorated by a single Al layer, namely monolayer AlB$_2$. Accordingly, similar superconducting transition temperature could be expected in honeycomb borophene on Al(111)., Comment: accepted for publication in Phys. Rev. B

An intact CuO$_2$ plane is widely believed to be a prerequisite for the high-$T_c$ superconductivity in cuprate superconductors. However, an exception may exist in the superconducting Ba$_2$CuO$_{3+\delta}$ materials where CuO chains play a more important role. From first-principles density functional theory calculations, we have studied the electronic and magnetic structures of Ba$_2$CuO$_{3+\delta}$. The stoichiometric Ba$_2$CuO$_3$ and Ba$_2$CuO$_4$ contain quasi-one-dimensional CuO chains and intact two-dimensional CuO$_2$ planes, respectively. In comparison with the nonmagnetic metal Ba$_2$CuO$_4$, Ba$_2$CuO$_3$ is found to be an antiferromagnetic (AFM) Mott insulator. It possesses a nearest-neighbor intra-chain antiferromagnetic (AFM) coupling and a weak inter-chain interaction, and its lowest unoccupied band and highest occupied band are contributed by Cu 3$d_{b^2-c^2}$-orbital (or $d_{x^2-y^2}$-orbital if we denote the $bc$-plane as the $xy$-plane) and O 2$p$-orbitals, respectively. Total energy calculations indicate that the oxygen vacancies in Ba$_2$CuO$_{3+\delta}$ prefer to reside in the planar sites rather than the apical oxygens in the CuO chains, in agreement with the experimental observation. Furthermore, we find that the magnetic frustrations or spin fluctuations can be effectively induced by moderate charge doping. This suggests that the superconducting pairing in oxygen-enriched Ba$_2$CuO$_{3+\delta}$ or oxygen-deficient Ba$_2$CuO$_{4-\delta}$ is likely to be mainly driven by the AFM fluctuations within CuO chains., Comment: 7 pages, 7 figures, 3 tables

Matrix product states (MPS), a tensor network designed for one-dimensional quantum systems, has been recently proposed for generative modeling of natural data (such as images) in terms of `Born machine'. However, the exponential decay of correlation in MPS restricts its representation power heavily for modeling complex data such as natural images. In this work, we push forward the effort of applying tensor networks to machine learning by employing the Tree Tensor Network (TTN) which exhibits balanced performance in expressibility and efficient training and sampling. We design the tree tensor network to utilize the 2-dimensional prior of the natural images and develop sweeping learning and sampling algorithms which can be efficiently implemented utilizing Graphical Processing Units (GPU). We apply our model to random binary patterns and the binary MNIST datasets of handwritten digits. We show that TTN is superior to MPS for generative modeling in keeping correlation of pixels in natural images, as well as giving better log-likelihood scores in standard datasets of handwritten digits. We also compare its performance with state-of-the-art generative models such as the Variational AutoEncoders, Restricted Boltzmann machines, and PixelCNN. Finally, we discuss the future development of Tensor Network States in machine learning problems.

Automated deception detection (ADD) from real-life videos is a challenging task. It specifically needs to address two problems: (1) Both face and body contain useful cues regarding whether a subject is deceptive. How to effectively fuse the two is thus key to the effectiveness of an ADD model. (2) Real-life deceptive samples are hard to collect; learning with limited training data thus challenges most deep learning based ADD models. In this work, both problems are addressed. Specifically, for face-body multimodal learning, a novel face-focused cross-stream network (FFCSN) is proposed. It differs significantly from the popular two-stream networks in that: (a) face detection is added into the spatial stream to capture the facial expressions explicitly, and (b) correlation learning is performed across the spatial and temporal streams for joint deep feature learning across both face and body. To address the training data scarcity problem, our FFCSN model is trained with both meta learning and adversarial learning. Extensive experiments show that our FFCSN model achieves state-of-the-art results. Further, the proposed FFCSN model as well as its robust training strategy are shown to be generally applicable to other human-centric video analysis tasks such as emotion recognition from user-generated videos.

Zero-shot learning (ZSL) aims to recognize a set of unseen classes without any training images. The standard approach to ZSL requires a set of training images annotated with seen class labels and a semantic descriptor for seen/unseen classes (attribute vector is the most widely used). Class label/attribute annotation is expensive; it thus severely limits the scalability of ZSL. In this paper, we define a new ZSL setting where only a few annotated images are collected from each seen class. This is clearly more challenging yet more realistic than the conventional ZSL setting. To overcome the resultant image-level attribute sparsity, we propose a novel inductive ZSL model termed sparse attribute propagation (SAP) by propagating attribute annotations to more unannotated images using sparse coding. This is followed by learning bidirectional projections between features and attributes for ZSL. An efficient solver is provided, together with rigorous theoretic algorithm analysis. With our SAP, we show that a ZSL training dataset can now be augmented by the abundant web images returned by image search engine, to further improve the model performance. Moreover, the general applicability of SAP is demonstrated on solving the social image annotation (SIA) problem. Extensive experiments show that our model achieves superior performance on both ZSL and SIA.

In this paper, a unified approach is presented to transfer learning that addresses several source and target domain label-space and annotation assumptions with a single model. It is particularly effective in handling a challenging case, where source and target label-spaces are disjoint, and outperforms alternatives in both unsupervised and semi-supervised settings. The key ingredient is a common representation termed Common Factorised Space. It is shared between source and target domains, and trained with an unsupervised factorisation loss and a graph-based loss. With a wide range of experiments, we demonstrate the flexibility, relevance and efficacy of our method, both in the challenging cases with disjoint label spaces, and in the more conventional cases such as unsupervised domain adaptation, where the source and target domains share the same label-sets., Comment: AAAI-19