Search

Quantum secret sharing (QSS) is one of the basic communication primitives in future quantum networks which addresses part of the basic cryptographic tasks of multiparty communication and computation. Nevertheless, it is a challenge to provide a practical QSS protocol with security against general attacks. A QSS protocol that balances security and practicality is still lacking. Here, we propose a QSS protocol with simple phase encoding of coherent states among three parties. Removing the requirement of impractical entangled resources and the need for phase randomization, our protocol can be implemented with accessible technology. We provide the finite-key analysis against coherent attacks and implement a proof-of-principle experiment to demonstrate our scheme's feasibility. Our scheme achieves a key rate of 85.3 bps under a 35 dB channel loss. Combined with security against general attacks and accessible technology, our protocol is a promising candidate for practical multiparty quantum communication networks., Comment: 10 pages, 5 figures, 3 tables, accepted by Sci. China-Phys. Mech. Astron

As Internet of Things devices become prevalent, using intrusion detection to protect IoT from malicious intrusions is of vital importance. However, the data scarcity of IoT hinders the effectiveness of traditional intrusion detection methods. To tackle this issue, in this paper, we propose the Adaptive Bi-Recommendation and Self-Improving Network (ABRSI) based on unsupervised heterogeneous domain adaptation (HDA). The ABRSI transfers enrich intrusion knowledge from a data-rich network intrusion source domain to facilitate effective intrusion detection for data-scarce IoT target domains. The ABRSI achieves fine-grained intrusion knowledge transfer via adaptive bi-recommendation matching. Matching the bi-recommendation interests of two recommender systems and the alignment of intrusion categories in the shared feature space form a mutual-benefit loop. Besides, the ABRSI uses a self-improving mechanism, autonomously improving the intrusion knowledge transfer from four ways. A hard pseudo label voting mechanism jointly considers recommender system decision and label relationship information to promote more accurate hard pseudo label assignment. To promote diversity and target data participation during intrusion knowledge transfer, target instances failing to be assigned with a hard pseudo label will be assigned with a probabilistic soft pseudo label, forming a hybrid pseudo-labelling strategy. Meanwhile, the ABRSI also makes soft pseudo-labels globally diverse and individually certain. Finally, an error knowledge learning mechanism is utilised to adversarially exploit factors that causes detection ambiguity and learns through both current and previous error knowledge, preventing error knowledge forgetfulness. Holistically, these mechanisms form the ABRSI model that boosts IoT intrusion detection accuracy via HDA-assisted intrusion knowledge transfer., Comment: Accepted by IEEE Internet of Things Journal

As the capacity of Solid-State Drives (SSDs) is constantly being optimised and boosted with gradually reduced cost, the SSD cluster is now widely deployed as part of the hybrid storage system in various scenarios such as cloud computing and big data processing. However, despite its rapid developments, the performance of the SSD cluster remains largely under-investigated, leaving its sub-optimal applications in reality. To address this issue, in this paper we conduct extensive empirical studies for a comprehensive understanding of the SSD cluster in diverse settings. To this end, we configure a real SSD cluster and gather the generated trace data based on some often-used benchmarks, then adopt analytical methods to analyse the performance of the SSD cluster with different configurations. In particular, regression models are built to provide better performance predictability under broader configurations, and the correlations between influential factors and performance metrics with respect to different numbers of nodes are investigated, which reveal the high scalability of the SSD cluster. Additionally, the cluster's network bandwidth is inspected to explain the performance bottleneck. Finally, the knowledge gained is summarised to benefit the SSD cluster deployment in practice., Comment: Accepted by Concurrency and Computation: Practice and Experience

Event cameras, such as dynamic vision sensors (DVS), are biologically inspired vision sensors that have advanced over conventional cameras in high dynamic range, low latency and low power consumption, showing great application potential in many fields. Event cameras are more sensitive to junction leakage current and photocurrent as they output differential signals, losing the smoothing function of the integral imaging process in the RGB camera. The logarithmic conversion further amplifies noise, especially in low-contrast conditions. Recently, researchers proposed a series of datasets and evaluation metrics but limitations remain: 1) the existing datasets are small in scale and insufficient in noise diversity, which cannot reflect the authentic working environments of event cameras; and 2) the existing denoising evaluation metrics are mostly referenced evaluation metrics, relying on APS information or manual annotation. To address the above issues, we construct a large-scale event denoising dataset (multilevel benchmark for event denoising, E-MLB) for the first time, which consists of 100 scenes, each with four noise levels, that is 12 times larger than the largest existing denoising dataset. We also propose the first nonreference event denoising metric, the event structural ratio (ESR), which measures the structural intensity of given events. ESR is inspired by the contrast metric, but is independent of the number of events and projection direction. Based on the proposed benchmark and ESR, we evaluate the most representative denoising algorithms, including classic and SOTA, and provide denoising baselines under various scenes and noise levels. The corresponding results and codes are available at https://github.com/KugaMaxx/cuke-emlb.

Due to the long coherence time and efficient manipulation, the surface electron (SE) provides a perfect two-dimensional platform for quantum computation and quantum simulation. In this work, a theoretical scheme to realize the controlled-NOT (CNOT) gate is proposed, where the two-qubit system is encoded on the four-level Rydberg structure of SE. The state transfer is achieved by a three-level structure with an intermediate level. By simultaneously driving the SE with two external electromagnetic fields, the dark state in the electromagnetically induced transparency (EIT) effect is exploited to suppress the population of the most dissipative state and increase the robustness against dissipation. The fidelity of the scheme is 0.9989 with experimentally achievable parameters.

Data-free quantization (DFQ) recovers the performance of quantized network (Q) without the original data, but generates the fake sample via a generator (G) by learning from full-precision network (P), which, however, is totally independent of Q, overlooking the adaptability of the knowledge from generated samples, i.e., informative or not to the learning process of Q, resulting into the overflow of generalization error. Building on this, several critical questions -- how to measure the sample adaptability to Q under varied bit-width scenarios? whether the largest adaptability is the best? how to generate the samples with adaptive adaptability to improve Q's generalization? To answer the above questions, in this paper, we propose an Adaptive Data-Free Quantization (AdaDFQ) method, which revisits DFQ from a zero-sum game perspective upon the sample adaptability between two players -- a generator and a quantized network. Following this viewpoint, we further define the disagreement and agreement samples to form two boundaries, where the margin is optimized to adaptively regulate the adaptability of generated samples to Q, so as to address the over-and-under fitting issues. Our AdaDFQ reveals: 1) the largest adaptability is NOT the best for sample generation to benefit Q's generalization; 2) the knowledge of the generated sample should not be informative to Q only, but also related to the category and distribution information of the training data for P. The theoretical and empirical analysis validate the advantages of AdaDFQ over the state-of-the-arts. Our code is available at https://github.com/hfutqian/AdaDFQ., Comment: 9 pages, 6 figures, Refined camera ready version for CVPR 2023

Estimating the 3DoF rotation from a single RGB image is an important yet challenging problem. Probabilistic rotation regression has raised more and more attention with the benefit of expressing uncertainty information along with the prediction. Though modeling noise using Gaussian-resembling Bingham distribution and matrix Fisher distribution is natural, they are shown to be sensitive to outliers for the nature of quadratic punishment to deviations. In this paper, we draw inspiration from multivariate Laplace distribution and propose a novel Rotation Laplace distribution on SO(3). Rotation Laplace distribution is robust to the disturbance of outliers and enforces much gradient to the low-error region, resulting in a better convergence. Our extensive experiments show that our proposed distribution achieves state-of-the-art performance for rotation regression tasks over both probabilistic and non-probabilistic baselines. Our project page is at https://pku-epic.github.io/RotationLaplace., Comment: ICLR 2023 Spotlight (notable-top-25%)

Data-free quantization (DFQ) recovers the performance of quantized network (Q) without accessing the real data, but generates the fake sample via a generator (G) by learning from full-precision network (P) instead. However, such sample generation process is totally independent of Q, specialized as failing to consider the adaptability of the generated samples, i.e., beneficial or adversarial, over the learning process of Q, resulting into non-ignorable performance loss. Building on this, several crucial questions -- how to measure and exploit the sample adaptability to Q under varied bit-width scenarios? how to generate the samples with desirable adaptability to benefit the quantized network? -- impel us to revisit DFQ. In this paper, we answer the above questions from a game-theory perspective to specialize DFQ as a zero-sum game between two players -- a generator and a quantized network, and further propose an Adaptability-aware Sample Generation (AdaSG) method. Technically, AdaSG reformulates DFQ as a dynamic maximization-vs-minimization game process anchored on the sample adaptability. The maximization process aims to generate the sample with desirable adaptability, such sample adaptability is further reduced by the minimization process after calibrating Q for performance recovery. The Balance Gap is defined to guide the stationarity of the game process to maximally benefit Q. The theoretical analysis and empirical studies verify the superiority of AdaSG over the state-of-the-arts. Our code is available at https://github.com/hfutqian/AdaSG., Comment: 9 pages, 7 figures, accepted by AAAI 2023

Text-to-image synthesis refers to generating visual-realistic and semantically consistent images from given textual descriptions. Previous approaches generate an initial low-resolution image and then refine it to be high-resolution. Despite the remarkable progress, these methods are limited in fully utilizing the given texts and could generate text-mismatched images, especially when the text description is complex. We propose a novel Fine-grained text-image Fusion based Generative Adversarial Networks, dubbed FF-GAN, which consists of two modules: Fine-grained text-image Fusion Block (FF-Block) and Global Semantic Refinement (GSR). The proposed FF-Block integrates an attention block and several convolution layers to effectively fuse the fine-grained word-context features into the corresponding visual features, in which the text information is fully used to refine the initial image with more details. And the GSR is proposed to improve the global semantic consistency between linguistic and visual features during the refinement process. Extensive experiments on CUB-200 and COCO datasets demonstrate the superiority of FF-GAN over other state-of-the-art approaches in generating images with semantic consistency to the given texts.Code is available at https://github.com/haoranhfut/FF-GAN., Comment: 13 pages, 8 figures, accepted by Chinese Journal of Electronics

On-device Deep Neural Network (DNN) inference consumes significant computing resources and development efforts. To alleviate that, we propose LUT-NN, the first system to empower inference by table lookup, to reduce inference cost. LUT-NN learns the typical features for each operator, named centroid, and precompute the results for these centroids to save in lookup tables. During inference, the results of the closest centroids with the inputs can be read directly from the table, as the approximated outputs without computations. LUT-NN integrates two major novel techniques: (1) differentiable centroid learning through backpropagation, which adapts three levels of approximation to minimize the accuracy impact by centroids; (2) table lookup inference execution, which comprehensively considers different levels of parallelism, memory access reduction, and dedicated hardware units for optimal performance. LUT-NN is evaluated on multiple real tasks, covering image and speech recognition, and nature language processing. Compared to related work, LUT-NN improves accuracy by 66% to 92%, achieving similar level with the original models. LUT-NN reduces the cost at all dimensions, including FLOPs ($\leq$ 16x), model size ($\leq$ 7x), latency ($\leq$ 6.8x), memory ($\leq$ 6.5x), and power ($\leq$ 41.7%).

This paper analyzes the convergence rate of a deep Galerkin method for the weak solution (DGMW) of second-order elliptic partial differential equations on $\mathbb{R}^d$ with Dirichlet, Neumann, and Robin boundary conditions, respectively. In DGMW, a deep neural network is applied to parametrize the PDE solution, and a second neural network is adopted to parametrize the test function in the traditional Galerkin formulation. By properly choosing the depth and width of these two networks in terms of the number of training samples $n$, it is shown that the convergence rate of DGMW is $\mathcal{O}(n^{-1/d})$, which is the first convergence result for weak solutions. The main idea of the proof is to divide the error of the DGMW into an approximation error and a statistical error. We derive an upper bound on the approximation error in the $H^{1}$ norm and bound the statistical error via Rademacher complexity., Comment: arXiv admin note: substantial text overlap with arXiv:2107.14478

Spatiotemporal learning, which aims at extracting spatiotemporal correlations from the collected spatiotemporal data, is a research hotspot in recent years. And considering the inherent graph structure of spatiotemporal data, recent works focus on capturing spatial dependencies by utilizing Graph Convolutional Networks (GCNs) to aggregate vertex features with the guidance of adjacency matrices. In this paper, with extensive and deep-going experiments, we comprehensively analyze existing spatiotemporal graph learning models and reveal that extracting adjacency matrices with carefully design strategies, which are viewed as the key of enhancing performance on graph learning, are largely ineffective. Meanwhile, based on these experiments, we also discover that the aggregation itself is more important than the way that how vertices are aggregated. With these preliminary, a novel efficient Graph-Free Spatial (GFS) learning module based on layer normalization for capturing spatial correlations in spatiotemporal graph learning. The proposed GFS module can be easily plugged into existing models for replacing all graph convolution components. Rigorous theoretical proof demonstrates that the time complexity of GFS is significantly better than that of graph convolution operation. Extensive experiments verify the superiority of GFS in both the perspectives of efficiency and learning effect in processing graph-structured data especially extreme large scale graph data.

Data scarcity hinders the usability of data-dependent algorithms when tackling IoT intrusion detection (IID). To address this, we utilise the data rich network intrusion detection (NID) domain to facilitate more accurate intrusion detection for IID domains. In this paper, a Geometric Graph Alignment (GGA) approach is leveraged to mask the geometric heterogeneities between domains for better intrusion knowledge transfer. Specifically, each intrusion domain is formulated as a graph where vertices and edges represent intrusion categories and category-wise interrelationships, respectively. The overall shape is preserved via a confused discriminator incapable to identify adjacency matrices between different intrusion domain graphs. A rotation avoidance mechanism and a centre point matching mechanism is used to avoid graph misalignment due to rotation and symmetry, respectively. Besides, category-wise semantic knowledge is transferred to act as vertex-level alignment. To exploit the target data, a pseudo-label election mechanism that jointly considers network prediction, geometric property and neighbourhood information is used to produce fine-grained pseudo-label assignment. Upon aligning the intrusion graphs geometrically from different granularities, the transferred intrusion knowledge can boost IID performance. Comprehensive experiments on several intrusion datasets demonstrate state-of-the-art performance of the GGA approach and validate the usefulness of GGA constituting components., Comment: Accepted by IEEE Internet of Things Journal

Structured light has attracted great interest in scientific and technical fields. Here, we demonstrate the first generation of structured air lasing in N2+ driven by 800 nm femtosecond laser pulses. By focusing a vortex pump beam at 800 nm in N2 gas, we generate a vortex superfluorescent radiation of N2+ at 391 nm, which carries the same photon orbital angular momentum as the pump beam. With the injection of a Gaussian seed beam at 391 nm, the coherent radiation is amplified, but the vorticity is unchanged. A new physical mechanism is revealed in the vortex N2+ superfluorescent radiation: the vortex pump beam transfers the spatial spiral phase into the N2+ gain medium, and the Gaussian seed beam picks up the spatial spiral phase and is then amplified into a vortex beam. Moreover, when we employ a pump beam with a cylindrical vector mode, the Gaussian seed beam is correspondingly amplified into a cylindrical vector beam. Surprisingly, the spatial polarization state of the amplified radiation is identical to that of the vector pump beam regardless of whether the Gaussian seed beam is linearly, elliptically, or circularly polarized. Solving three-dimensional coupled wave equations, we show how a Gaussian beam becomes a cylindrical vector beam in a cylindrically symmetric gain medium. This study provides a novel approach to generating structured light via N2+ air lasing., Comment: 18 pages, 5 figures, 3 equations

The peculiar velocities of dark matter halos are crucial to study many issues in cosmology and galaxy evolution. In this study, by using the state-of-the-art deep learning technique, a UNet-based neural network, we propose to reconstruct the peculiar velocity field from the redshift-space distribution of dark matter halos. Through a point-to-point comparison and examination of various statistical properties, we demonstrate that, the reconstructed velocity field is in good agreement with the ground truth. The power spectra of various velocity field components, including velocity magnitude, divergence and vorticity, can be successfully recovered when $k\lesssim 1.1$ $h/\rm Mpc$ (the Nyquist frequency of the simulations) at about 80% accuracy. This approach is very promising and presents an alternative method to correct the redshift-space distortions using the measured 3D spatial information of halos. Additionally, for the reconstruction of the momentum field of halos, UNet achieves similar good results. Hence the applications in various aspects of cosmology are very broad, such as correcting redshift errors and improving measurements in the structure of the cosmic web, the kinetic Sunyaev-Zel'dovich effect, BAO reconstruction, etc., Comment: 15 pages, 13 figures, 4 tables

Deep Neural Networks (DNNs) suffer from domain shift when the test dataset follows a distribution different from the training dataset. Domain generalization aims to tackle this issue by learning a model that can generalize to unseen domains. In this paper, we propose a new approach that aims to explicitly remove domain-specific features for domain generalization. Following this approach, we propose a novel framework called Learning and Removing Domain-specific features for Generalization (LRDG) that learns a domain-invariant model by tactically removing domain-specific features from the input images. Specifically, we design a classifier to effectively learn the domain-specific features for each source domain, respectively. We then develop an encoder-decoder network to map each input image into a new image space where the learned domain-specific features are removed. With the images output by the encoder-decoder network, another classifier is designed to learn the domain-invariant features to conduct image classification. Extensive experiments demonstrate that our framework achieves superior performance compared with state-of-the-art methods., Comment: 13 pages, 3 figures

Dermatological diseases are among the most common disorders worldwide. This paper presents the first study of the interpretability and imbalanced semi-supervised learning of the multiclass intelligent skin diagnosis framework (ISDL) using 58,457 skin images with 10,857 unlabeled samples. Pseudo-labelled samples from minority classes have a higher probability at each iteration of class-rebalancing self-training, thereby promoting the utilization of unlabeled samples to solve the class imbalance problem. Our ISDL achieved a promising performance with an accuracy of 0.979, sensitivity of 0.975, specificity of 0.973, macro-F1 score of 0.974 and area under the receiver operating characteristic curve (AUC) of 0.999 for multi-label skin disease classification. The Shapley Additive explanation (SHAP) method is combined with our ISDL to explain how the deep learning model makes predictions. This finding is consistent with the clinical diagnosis. We also proposed a sampling distribution optimisation strategy to select pseudo-labelled samples in a more effective manner using ISDLplus. Furthermore, it has the potential to relieve the pressure placed on professional doctors, as well as help with practical issues associated with a shortage of such doctors in rural areas.

In this paper, we allocate IoT devices as resources for smart services with time-constrained resource requirements. The allocation method named as BRAD can work under multiple resource scenarios with diverse resource richnesses, availabilities and costs, such as the intelligent healthcare system deployed by Harbin Institute of Technology (HIT-IHC). The allocation aims for bimetric-balancing under the multi-scenario case, i.e., the profit and cost associated with service satisfaction are jointly optimised and balanced wisely. Besides, we abstract IoT devices as digital objects (DO) to make them easier to interact with during resource allocation. Considering that the problem is NP-Hard and the optimisation objective is not differentiable, we utilise Grey Wolf Optimisation (GWO) algorithm as the model optimiser. Specifically, we tackle the deficiencies of GWO and significantly improve its performance by introducing three new mechanisms to form the BRAD-GWA algorithm. Comprehensive experiments are conducted on realistic HIT-IHC IoT testbeds and several algorithms are compared, including the allocation method originally used by HIT-IHC system to verify the effectiveness of the BRAD-GWA. The BRAD-GWA achieves a 3.14 times and 29.6% objective reduction compared with the HIT-IHC and the original GWO algorithm, respectively., Comment: Accepted by IEEE HPCC 2022

Few-shot segmentation (FSS) aims to segment objects of unseen classes given only a few annotated support images. Most existing methods simply stitch query features with independent support prototypes and segment the query image by feeding the mixed features to a decoder. Although significant improvements have been achieved, existing methods are still face class biases due to class variants and background confusion. In this paper, we propose a joint framework that combines more valuable class-aware and class-agnostic alignment guidance to facilitate the segmentation. Specifically, we design a hybrid alignment module which establishes multi-scale query-support correspondences to mine the most relevant class-aware information for each query image from the corresponding support features. In addition, we explore utilizing base-classes knowledge to generate class-agnostic prior mask which makes a distinction between real background and foreground by highlighting all object regions, especially those of unseen classes. By jointly aggregating class-aware and class-agnostic alignment guidance, better segmentation performances are obtained on query images. Extensive experiments on PASCAL-$5^i$ and COCO-$20^i$ datasets demonstrate that our proposed joint framework performs better, especially on the 1-shot setting.

In this paper, we propose a Joint Semantic Transfer Network (JSTN) towards effective intrusion detection for large-scale scarcely labelled IoT domain. As a multi-source heterogeneous domain adaptation (MS-HDA) method, the JSTN integrates a knowledge rich network intrusion (NI) domain and another small-scale IoT intrusion (II) domain as source domains, and preserves intrinsic semantic properties to assist target II domain intrusion detection. The JSTN jointly transfers the following three semantics to learn a domain-invariant and discriminative feature representation. The scenario semantic endows source NI and II domain with characteristics from each other to ease the knowledge transfer process via a confused domain discriminator and categorical distribution knowledge preservation. It also reduces the source-target discrepancy to make the shared feature space domain-invariant. Meanwhile, the weighted implicit semantic transfer boosts discriminability via a fine-grained knowledge preservation, which transfers the source categorical distribution to the target domain. The source-target divergence guides the importance weighting during knowledge preservation to reflect the degree of knowledge learning. Additionally, the hierarchical explicit semantic alignment performs centroid-level and representative-level alignment with the help of a geometric similarity-aware pseudo-label refiner, which exploits the value of unlabelled target II domain and explicitly aligns feature representations from a global and local perspective in a concentrated manner. Comprehensive experiments on various tasks verify the superiority of the JSTN against state-of-the-art comparing methods, on average a 10.3% of accuracy boost is achieved. The statistical soundness of each constituting component and the computational efficiency are also verified., Comment: Accepted by IEEE Internet of Things Journal

The dispersion measure(DM) of fast radio burst encodes important information such as its distance, properties of intervening medium. Based on simulations in the Illustris and IllustrisTNG projects, we analyze the DM of FRBs contributed by the interstellar medium and circumgalactic medium in the hosts, $\rm{DM_{host}}$. We explore two population models - tracing the star formation rate (SFR), and the stellar mass, i.e. young and old progenitors respectively. The distribution of $\rm{DM_{host}}$ shows significant differences at $z=0$ between two populations: the stellar mass model exhibits an excess at the low DM end with respect to the SFR model. The SFR (stellar mass) model has a median value of $\rm{DM_{host}}$=179 (63) $\rm{pc\, cm^{-3}}$ for galaxies with $M_*=10^{8-13}\,M_{\odot}$ in the TNG100-1. Galaxies in the Illustris-1 have a much smaller $\rm{DM_{host}}$. The distributions of $\rm{DM_{host}}$ deviate from log-normal function for both models. Furthermore, two populations differ moderately in the spatial offset from host galaxy's center, in the stellar mass function of hosts. $\rm{DM_{host}}$ increases with the stellar mass of hosts when $M_*<10^{10.5}\,M_{\odot}$, and fluctuate at higher mass. At $0

We study the stochastic contextual bandit with knapsacks (CBwK) problem, where each action, taken upon a context, not only leads to a random reward but also costs a random resource consumption in a vector form. The challenge is to maximize the total reward without violating the budget for each resource. We study this problem under a general realizability setting where the expected reward and expected cost are functions of contexts and actions in some given general function classes $\mathcal{F}$ and $\mathcal{G}$, respectively. Existing works on CBwK are restricted to the linear function class since they use UCB-type algorithms, which heavily rely on the linear form and thus are difficult to extend to general function classes. Motivated by online regression oracles that have been successfully applied to contextual bandits, we propose the first universal and optimal algorithmic framework for CBwK by reducing it to online regression. We also establish the lower regret bound to show the optimality of our algorithm for a variety of function classes., Comment: AISTATS2023

With the broad growth of video capturing devices and applications on the web, it is more demanding to provide desired video content for users efficiently. Video summarization facilitates quickly grasping video content by creating a compact summary of videos. Much effort has been devoted to automatic video summarization, and various problem settings and approaches have been proposed. Our goal is to provide an overview of this field. This survey covers early studies as well as recent approaches which take advantage of deep learning techniques. We describe video summarization approaches and their underlying concepts. We also discuss benchmarks and evaluations. We overview how prior work addressed evaluation and detail the pros and cons of the evaluation protocols. Last but not least, we discuss open challenges in this field., Comment: 53 pages

Ab-initio crystal structure prediction depends on accurate calculation of the energies of competing structures. Many DFT codes are available that utilize different approaches to solve the Kohn-Sham equation. We evaluate the consistency of three software packages (WIEN2k, VASP and MuST) that utilize three different methods (FL-APW, plane-wave pseudopotential and the KKR-Green's Function methods) by comparing the relative total energies obtained for a set of BCC and FCC binary metallic alloys. We focus on the impact of choices such as muffin-tin $vs.$ full-potential, angular momentum cutoff and other important KKR parameters. Different alloy systems prove more or less sensitive to these choices, and we explain the differences through study of the angular variation of their potentials. Our results can provide guidance in the application of KKR as a total energy method for structure prediction., Comment: 18 pages, 9 figures

We consider a new problem of few-shot learning of compact models. Meta-learning is a popular approach for few-shot learning. Previous work in meta-learning typically assumes that the model architecture during meta-training is the same as the model architecture used for final deployment. In this paper, we challenge this basic assumption. For final deployment, we often need the model to be small. But small models usually do not have enough capacity to effectively adapt to new tasks. In the mean time, we often have access to the large dataset and extensive computing power during meta-training since meta-training is typically performed on a server. In this paper, we propose task-specific meta distillation that simultaneously learns two models in meta-learning: a large teacher model and a small student model. These two models are jointly learned during meta-training. Given a new task during meta-testing, the teacher model is first adapted to this task, then the adapted teacher model is used to guide the adaptation of the student model. The adapted student model is used for final deployment. We demonstrate the effectiveness of our approach in few-shot image classification using model-agnostic meta-learning (MAML). Our proposed method outperforms other alternatives on several benchmark datasets., Comment: This paper has been accepted by WACV'2023

We investigate the effect of channel aging on multi-cell massive multiple-input multiple-output (MIMO) vehicular networks in a generic non-isotropic scattering environment. Based on the single cluster scattering assumption and the von Mises distribution assumptions of the scatterers' angles, an aging channel model is established to capture the joint effect of spatial and temporal correlations resulting from different angular spread conditions in various application scenarios. Expressions of the user uplink transmission spectral efficiency (SE) are derived for maximum ratio (MR) and minimum mean square error (MMSE) combining. Through numerical studies, the area spectral efficiency (ASE) performance of the network is evaluated in freeway and urban Manhattan road grid scenarios, and easy-to-use empirical models for the optimal transmission block size for ASE maximization are obtained for the evaluated scenarios.

Deep learning applications are being transferred from the cloud to edge with the rapid development of embedded computing systems. In order to achieve higher energy efficiency with the limited resource budget, neural networks(NNs) must be carefully designed in two steps, the architecture design and the quantization policy choice. Neural Architecture Search(NAS) and Quantization have been proposed separately when deploying NNs onto embedded devices. However, taking the two steps individually is time-consuming and leads to a sub-optimal final deployment. To this end, we propose a novel neural network design framework called Hardware-aware Quantized Neural Architecture Search(HQNAS) framework which combines the NAS and Quantization together in a very efficient manner using weight-sharing and bit-sharing. It takes only 4 GPU hours to discover an outstanding NN policy on CIFAR10. It also takes only %10 GPU time to generate a comparable model on Imagenet compared to the traditional NAS method with 1.8x decrease of latency and a negligible accuracy loss of only 0.7%. Besides, our method can be adapted in a lifelong situation where the neural network needs to evolve occasionally due to changes of local data, environment and user preference.

Due to user privacy and regulatory restrictions, federate learning (FL) is proposed as a distributed learning framework for training deep neural networks (DNN) on decentralized data clients. Recent advancements in FL have applied Neural Architecture Search (NAS) to replace the predefined one-size-fit-all DNN model, which is not optimal for all tasks of various data distributions, with searchable DNN architectures. However, previous methods suffer from expensive communication cost rasied by frequent large model parameters transmission between the server and clients. Such difficulty is further amplified when combining NAS algorithms, which commonly require prohibitive computation and enormous model storage. Towards this end, we propose FAQS, an efficient personalized FL-NAS-Quantization framework to reduce the communication cost with three features: weight-sharing super kernels, bit-sharing quantization and masked transmission. FAQS has an affordable search time and demands very limited size of transmitted messages at each round. By setting different personlized pareto function loss on local clients, FAQS can yield heterogeneous hardware-aware models for various user preferences. Experimental results show that FAQS achieves average reduction of 1.58x in communication bandwith per round compared with normal FL framework and 4.51x compared with FL+NAS framwork.

Ultrafast light sources have become an indispensable tool to access and understand transient phenomenon in material science. However, a simple and easy-to-implement method for harmonic selection, with high transmission efficiency and pulse duration conservation, is still a challenge. Here we showcase and compare two approaches for selecting the desired harmonic from a high harmonic generation source while achieving the above goals. The first approach is the combination of extreme ultraviolet spherical mirrors with transmission filters and the second approach uses a normal-incidence spherical grating. Both solutions target time- and angle-resolved photoemission spectroscopy with photon energies in the 10-20 eV range but are relevant for other experimental techniques as well. The two approaches for harmonic selection are characterized in terms of focusing quality, efficiency, and temporal broadening. It is demonstrated that a focusing grating is able to provide much higher transmission as compared to the mirror+filter approach (3.3 times higher for 10.8 eV and 12.9 times higher for 18.1 eV), with only a slight temporal broadening (6.8% increase) and a somewhat larger spot size (~30% increase). Overall, our study establishes an experimental perspective on the trade-off between a single grating normal incidence monochromator design and the use of filters. As such, it provides a basis for selecting the most appropriate approach in various fields where an easy-to-implement harmonic selection from high harmonic generation is needed.

Few-shot object detection (FSOD) is to detect objects with a few examples. However, existing FSOD methods do not consider hierarchical fine-grained category structures of objects that exist widely in real life. For example, animals are taxonomically classified into orders, families, genera and species etc. In this paper, we propose and solve a new problem called hierarchical few-shot object detection (Hi-FSOD), which aims to detect objects with hierarchical categories in the FSOD paradigm. To this end, on the one hand, we build the first large-scale and high-quality Hi-FSOD benchmark dataset HiFSOD-Bird, which contains 176,350 wild-bird images falling to 1,432 categories. All the categories are organized into a 4-level taxonomy, consisting of 32 orders, 132 families, 572 genera and 1,432 species. On the other hand, we propose the first Hi-FSOD method HiCLPL, where a hierarchical contrastive learning approach is developed to constrain the feature space so that the feature distribution of objects is consistent with the hierarchical taxonomy and the model's generalization power is strengthened. Meanwhile, a probabilistic loss is designed to enable the child nodes to correct the classification errors of their parent nodes in the taxonomy. Extensive experiments on the benchmark dataset HiFSOD-Bird show that our method HiCLPL outperforms the existing FSOD methods., Comment: Accepted by ACM MM 2022

In this paper, we tackle the problem of domain shift. Most existing methods perform training on multiple source domains using a single model, and the same trained model is used on all unseen target domains. Such solutions are sub-optimal as each target domain exhibits its own specialty, which is not adapted. Furthermore, expecting single-model training to learn extensive knowledge from multiple source domains is counterintuitive. The model is more biased toward learning only domain-invariant features and may result in negative knowledge transfer. In this work, we propose a novel framework for unsupervised test-time adaptation, which is formulated as a knowledge distillation process to address domain shift. Specifically, we incorporate Mixture-of-Experts (MoE) as teachers, where each expert is separately trained on different source domains to maximize their specialty. Given a test-time target domain, a small set of unlabeled data is sampled to query the knowledge from MoE. As the source domains are correlated to the target domains, a transformer-based aggregator then combines the domain knowledge by examining the interconnection among them. The output is treated as a supervision signal to adapt a student prediction network toward the target domain. We further employ meta-learning to enforce the aggregator to distill positive knowledge and the student network to achieve fast adaptation. Extensive experiments demonstrate that the proposed method outperforms the state-of-the-art and validates the effectiveness of each proposed component. Our code is available at https://github.com/n3il666/Meta-DMoE., Comment: Accepted at NeurIPS2022

In this work, we present the winning solution for ORBIT Few-Shot Video Object Recognition Challenge 2022. Built upon the ProtoNet baseline, the performance of our method is improved with three effective techniques. These techniques include the embedding adaptation, the uniform video clip sampler and the invalid frame detection. In addition, we re-factor and re-implement the official codebase to encourage modularity, compatibility and improved performance. Our implementation accelerates the data loading in both training and testing., Comment: Winner of ORBIT Challenge 2022

Session-based recommendation (SBR) systems aim to utilize the user's short-term behavior sequence to predict the next item without the detailed user profile. Most recent works try to model the user preference by treating the sessions as between-item transition graphs and utilize various graph neural networks (GNNs) to encode the representations of pair-wise relations among items and their neighbors. Some of the existing GNN-based models mainly focus on aggregating information from the view of spatial graph structure, which ignores the temporal relations within neighbors of an item during message passing and the information loss results in a sub-optimal problem. Other works embrace this challenge by incorporating additional temporal information but lack sufficient interaction between the spatial and temporal patterns. To address this issue, inspired by the uniformity and alignment properties of contrastive learning techniques, we propose a novel framework called Session-based Recommendation with Spatio-Temporal Contrastive Learning Enhanced GNNs (RESTC). The idea is to supplement the GNN-based main supervised recommendation task with the temporal representation via an auxiliary cross-view contrastive learning mechanism. Furthermore, a novel global collaborative filtering graph (CFG) embedding is leveraged to enhance the spatial view in the main task. Extensive experiments demonstrate the significant performance of RESTC compared with the state-of-the-art baselines e.g., with an improvement as much as 27.08% gain on HR@20 and 20.10% gain on MRR@20., Comment: Accepted by ACM Transaction on Information Systems (ACM TOIS)

We present a new magnetometry method integrating an ensemble of nitrogen-vacancy (NV) centers in a single-crystal diamond with an extended dynamic range for monitoring the fast changing magnetic-field. The NV-center spin resonance frequency is tracked using a closed-loop frequency locked technique with fast frequency hopping to achieve a 10 kHz measurement bandwidth, thus, allowing for the detection of fast changing magnetic signals up to 0.723 T/s.This technique exhibits an extended dynamic range subjected to the working bandwidth of the microwave source. This extended dynamic range can reach up to 4.3 mT, which is 86 times broader than the intrinsic dynamic range. The essential components for NV spin control and signal processing such as signal generation, microwave frequency control, data processing and readout are integrated in a board-level system. With this platform, we demonstrate broadband magnetometry with an optimized sensitivity of 4.2 nT-Hz-1/2. This magnetometry method has the potential to be implemented in a multichannel frequency locked vector magnetometer suitable for a wide range of practical applications such as magnetocardiography and high-precision current sensors., Comment: 18 pages, 9 figures

Image inpainting has achieved remarkable progress and inspired abundant methods, where the critical bottleneck is identified as how to fulfill the high-frequency structure and low-frequency texture information on the masked regions with semantics. To this end, deep models exhibit powerful superiority to capture them, yet constrained on the local spatial regions. In this paper, we delve globally into texture and structure information to well capture the semantics for image inpainting. As opposed to the existing arts trapped on the independent local patches, the texture information of each patch is reconstructed from all other patches across the whole image, to match the coarsely filled information, specially the structure information over the masked regions. Unlike the current decoder-only transformer within the pixel level for image inpainting, our model adopts the transformer pipeline paired with both encoder and decoder. On one hand, the encoder captures the texture semantic correlations of all patches across image via self-attention module. On the other hand, an adaptive patch vocabulary is dynamically established in the decoder for the filled patches over the masked regions. Building on this, a structure-texture matching attention module anchored on the known regions comes up to marry the best of these two worlds for progressive inpainting via a probabilistic diffusion process. Our model is orthogonal to the fashionable arts, such as Convolutional Neural Networks (CNNs), Attention and Transformer model, from the perspective of texture and structure information for image inpainting. The extensive experiments over the benchmarks validate its superiority. Our code is available at https://github.com/htyjers/DGTS-Inpainting., Comment: 9 pages, 10 figures, accepted by ACM Multimedia 2022

In addition to the topologically protected linear dispersion, a band-bending-confined two-dimensional electron gas with tunable Rashba spin-splitting (RSS) was found to coexist with the topological surface states on the surface of topological insulators (TIs). Here, we report the observation of large bilinear magnetoresistance (BMR) in Bi2Se3 films decorated with transition metal atoms. The magnitude of the BMR sensitively depends on the type and amount of atoms deposited, with a maximum achieved value close to those of strong Rashba semiconductors. Our first-principles calculations reproduce the quantum well states and reveal sizable RSS in all Bi2Se3 heterostructures with broken inversion symmetry. Our results show that charge-spin interconversion through RSS states in TIs can be fine-tuned through surface atom deposition and easily detected via BMR for potential spintronic applications., Comment: Typos in Fig. 4 caption and Ref. [27] were corrected in this version

Online Knowledge Distillation (OKD) improves the involved models by reciprocally exploiting the difference between teacher and student. Several crucial bottlenecks over the gap between them -- e.g., Why and when does a large gap harm the performance, especially for student? How to quantify the gap between teacher and student? -- have received limited formal study. In this paper, we propose Switchable Online Knowledge Distillation (SwitOKD), to answer these questions. Instead of focusing on the accuracy gap at test phase by the existing arts, the core idea of SwitOKD is to adaptively calibrate the gap at training phase, namely distillation gap, via a switching strategy between two modes -- expert mode (pause the teacher while keep the student learning) and learning mode (restart the teacher). To possess an appropriate distillation gap, we further devise an adaptive switching threshold, which provides a formal criterion as to when to switch to learning mode or expert mode, and thus improves the student's performance. Meanwhile, the teacher benefits from our adaptive switching threshold and keeps basically on a par with other online arts. We further extend SwitOKD to multiple networks with two basis topologies. Finally, extensive experiments and analysis validate the merits of SwitOKD for classification over the state-of-the-arts. Our code is available at https://github.com/hfutqian/SwitOKD., Comment: 16 pages, 7 figures, accepted by ECCV 2022

Solar batteries capable of harvesting sunlight and storing solar energy present an attractive vista to transition our energy infrastructure into a sustainable future. We present an integrated, fully earth-abundant solar battery design based on a bifunctional (light absorbing and charge storing) carbon nitride (K-PHI) photoanode, combined with all-organic hole transfer and storage materials. An internal ladder-type hole transfer cascade via a hole transport layer is used to shuttle the photogenerated holes to the PEDOT:PSS cathode. This concept differs from previous designs such as light-assisted battery schemes or photocapacitors and allows charging with only light or light-assisted during electrical charging and discharging, thus substantially increasing the energy output of the cell. Compared to battery operation in the dark, light-assisted (dis)charging increases charge output by 243 %, thereby increasing the electric coulombic efficiency from 68.3 % in the dark to 231 %, leading to energy improvements of 94.1 % under illumination., Comment: manuscript and supplementary information

The electronic structure and the physical properties of quantum materials can be significantly altered by charge carrier doping and magnetic state transition. Here we report a discovery of a giant and reversible electronic structure evolution with doping in a magnetic topological material. By performing high-resolution angle-resolved photoemission measurements on EuCd2As2,we found that a huge amount of hole doping can be introduced into the sample surface due to surface absorption. The electronic structure exhibits a dramatic change with the hole doping which can not be described by a rigid band shift. Prominent band splitting is observed at high doping which corresponds to a doping-induced magnetic transition at low temperature (below -15 K) from an antiferromagnetic state to a ferromagnetic state. These results have established a detailed electronic phase diagram of EuCd2As2 where the electronic structure and the magnetic structure change systematically and dramatically with the doping level. They further suggest that the transport, magnetic and topological properties of EuCd2As2 can be greatly modified by doping. These work will stimulate further investigations to explore for new phenomena and properties in doping this magnetic topological material., Comment: 22 pages,5 figures

With photonics, the quantum computational advantage has been demonstrated on the task of boson sampling. Next, developing quantum-enhanced approaches for practical problems becomes one of the top priorities for photonic systems. Quantum walks are powerful kernels for developing new and useful quantum algorithms. Here we realize large-scale quantum walks using a fully programmable photonic quantum computing system. The system integrates a silicon quantum photonic chip, enabling the simulation of quantum walk dynamics on graphs with up to 400 vertices and possessing full programmability over quantum walk parameters, including the particle property, initial state, graph structure, and evolution time. In the 400-dimensional Hilbert space, the average fidelity of random entangled quantum states after the whole on-chip circuit evolution reaches as high as 94.29$\pm$1.28$\%$. With the system, we demonstrated exponentially faster hitting and quadratically faster mixing performance of quantum walks over classical random walks, achieving more than two orders of magnitude of enhancement in the experimental hitting efficiency and almost half of the reduction in the experimental evolution time for mixing. We utilize the system to implement a series of quantum applications, including measuring the centrality of scale-free networks, searching targets on Erd\"{o}s-R\'{e}nyi networks, distinguishing non-isomorphic graph pairs, and simulating the topological phase of higher-order topological insulators. Our work shows one feasible path for quantum photonics to address applications of practical interests in the near future.

Given a spectrally sparse signal $\mathbf{y} = \sum_{i=1}^s x_i\mathbf{f}(\tau_i) \in \mathbb{C}^{2n+1}$ consisting of $s$ complex sinusoids, we consider the super-resolution problem, which is about estimating frequency components $\{\tau_i\}_{i=1}^s$ of $\mathbf y$. We consider the OMP-type algorithms for super-resolution, which is more efficient than other approaches based on Semi-Definite Programming. Our analysis shows that a two-stage algorithm with OMP initialization can recover frequency components under the separation condition $n\Delta \gtrsim \text{dyn}(\mathbf{x})$ and the dependency on $\text{dyn}(\mathbf{x})$ is inevitable for the vanilla OMP algorithm. We further show that the Sliding-OMP algorithm, a variant of the OMP algorithm with an additional refinement step at each iteration, is provable to recover $\{\tau_i\}_{i=1}^s$ under the separation condition $n\Delta \geq c$. Moreover, our result can be extended to an incomplete measurement model with $O( s^2\log n)$ measurements., Comment: Add numerical experiments

Spatiotemporal forecasting is an imperative topic in data science due to its diverse and critical applications in smart cities. Existing works mostly perform consecutive predictions of following steps with observations completely and continuously obtained, where nearest observations can be exploited as key knowledge for instantaneous status estimation. However, the practical issues of early activity planning and sensor failures elicit a brand-new task, i.e., non-consecutive forecasting. In this paper, we define spatiotemporal learning systems with missing observation as Grey Spatiotemporal Systems (G2S) and propose a Factor-Decoupled learning framework for G2S (FDG2S), where the core idea is to hierarchically decouple multi-level factors and enable both flexible aggregations and disentangled uncertainty estimations. Firstly, to compensate for missing observations, a generic semantic-neighboring sequence sampling is devised, which selects representative sequences to capture both periodical regularity and instantaneous variations. Secondly, we turn the predictions of non-consecutive statuses into inferring statuses under expected combined exogenous factors. In particular, a factor-decoupled aggregation scheme is proposed to decouple factor-induced predictive intensity and region-wise proximity by two energy functions of conditional random field. To infer region-wise proximity under flexible factor-wise combinations and enable dynamic neighborhood aggregations, we further disentangle compounded influences of exogenous factors on region-wise proximity and learn to aggregate them. Given the inherent incompleteness and critical applications of G2S, a DisEntangled Uncertainty Quantification is put forward, to identify two types of uncertainty for reliability guarantees and model interpretations., Comment: 13 pages, 6 figures and 4 tables

Time series underwent the transition from statistics to deep learning, as did many other machine learning fields. Although it appears that the accuracy has been increasing as the model is updated in a number of publicly available datasets, it typically only increases the scale by several times in exchange for a slight difference in accuracy. Through this experiment, we point out a different line of thinking, time series, especially long-term forecasting, may differ from other fields. It is not necessary to use extensive and complex models to grasp all aspects of time series, but to use pure models to grasp the core rules of time series changes. With this simple but effective idea, we created PureTS, a network with three pure linear layers that achieved state-of-the-art in 80% of the long sequence prediction tasks while being nearly the lightest model and having the fastest running speed. On this basis, we discuss the potential of pure linear layers in both phenomena and essence. The ability to understand the core law contributes to the high precision of long-distance prediction, and reasonable fluctuation prevents it from distorting the curve in multi-step prediction like mainstream deep learning models, which is summarized as a pure linear neural network that avoids over-fluctuating. Finally, we suggest the fundamental design standards for lightweight long-step time series tasks: input and output should try to have the same dimension, and the structure avoids fragmentation and complex operations., Comment: To be submitted to the conference

Quality feature representation is key to instance image retrieval. To attain it, existing methods usually resort to a deep model pre-trained on benchmark datasets or even fine-tune the model with a task-dependent labelled auxiliary dataset. Although achieving promising results, this approach is restricted by two issues: 1) the domain gap between benchmark datasets and the dataset of a given retrieval task; 2) the required auxiliary dataset cannot be readily obtained. In light of this situation, this work looks into a different approach which has not been well investigated for instance image retrieval previously: {can we learn feature representation \textit{specific to} a given retrieval task in order to achieve excellent retrieval?} Our finding is encouraging. By adding an object proposal generator to generate image regions for self-supervised learning, the investigated approach can successfully learn feature representation specific to a given dataset for retrieval. This representation can be made even more effective by boosting it with image similarity information mined from the dataset. As experimentally validated, such a simple ``self-supervised learning + self-boosting'' approach can well compete with the relevant state-of-the-art retrieval methods. Ablation study is conducted to show the appealing properties of this approach and its limitation on generalisation across datasets.

To simply and effectively realize the trajectory tracking control of a bionic joint actuated by a single pneumatic artificial muscle (PAM), a cascaded control strategy is proposed based on the robust modeling method. Firstly, the relationship between the input voltage of the proportional directional control valve and the inner driving pressure of PAM is expressed as a nonlinear model analytically. Secondly, the nonlinear relationship between the driving pressure input of PAM and the angular position output of the bionic joint is described as a second-order linear time-invariant model (LTI) accompanied by parametric perturbations, equivalently, and then the parameters of the model are identified by the robust modeling method. Then, a hybrid model is established based on the two models (the nonlinear model and the LTI model) and corresponding to it, a cascaded controller is developed, the outer loop of which is an H-infinite controller for the angular position tracking designed by loop-shaping design procedure (LSDP) and the inner loop is a nonlinear controller based on the feedback linearization theory for the PAM driving pressure control. Finally, the experiment is accomplished within the joint rotation range of 90 degrees and with the working frequency upper bound of 1.25 rad/s. And the joint with the developed cascaded controller tracks given reference trajectories with steady-state errors smaller than 2%. Results show that the trajectory tracking control of a highly nonlinear system is highly efficient using the proposed strategy in the case of relatively low work frequency., Comment: 7 pages, 16 figures, journal paper

With the fast growing quantity of data generated by smart devices and the exponential surge of processing demand in the Internet of Things (IoT) era, the resource-rich cloud centres have been utilised to tackle these challenges. To relieve the burden on cloud centres, edge-cloud computation offloading becomes a promising solution since shortening the proximity between the data source and the computation by offloading computation tasks from the cloud to edge devices can improve performance and Quality of Service (QoS). Several optimisation models of edge-cloud computation offloading have been proposed that take computation costs and heterogeneous communication costs into account. However, several important factors are not jointly considered, such as heterogeneities of tasks, load balancing among nodes and the profit yielded by computation tasks, which lead to the profit and cost-oriented computation offloading optimisation model PECCO proposed in this paper. Considering that the model is hard in nature and the optimisation objective is not differentiable, we propose an improved Moth-flame optimiser PECCO-MFI which addresses some deficiencies of the original Moth-flame Optimiser and integrate it under the edge-cloud environment. Comprehensive experiments are conducted to verify the superior performance of the proposed method when optimising the proposed task offloading model under the edge-cloud environment., Comment: Accepted by Concurrency and Computation: Practice and Experience

A large volume of remote sensing (RS) data has been generated with the deployment of satellite technologies. The data facilitates research in ecological monitoring, land management and desertification, etc. The characteristics of RS data (e.g., enormous volume, large single-file size and demanding requirement of fault tolerance) make the Hadoop Distributed File System (HDFS) an ideal choice for RS data storage as it is efficient, scalable and equipped with a data replication mechanism for failure resilience. To use RS data, one of the most important techniques is geospatial indexing. However, the large data volume makes it time-consuming to efficiently construct and leverage. Considering that most modern geospatial data centres are equipped with HDFS-based big data processing infrastructures, deploying multiple geospatial indices becomes natural to optimise the efficacy. Moreover, because of the reliability introduced by high-quality hardware and the infrequently modified property of the RS data, the use of multi-indexing will not cause large overhead. Therefore, we design a framework called Multi-IndeXing-RS (MIX-RS) that unifies the multi-indexing mechanism on top of the HDFS with data replication enabled for both fault tolerance and geospatial indexing efficiency. Given the fault tolerance provided by the HDFS, RS data is structurally stored inside for faster geospatial indexing. Additionally, multi-indexing enhances efficiency. The proposed technique naturally sits on top of the HDFS to form a holistic framework without incurring severe overhead or sophisticated system implementation efforts. The MIX-RS framework is implemented and evaluated using real remote sensing data provided by the Chinese Academy of Sciences, demonstrating excellent geospatial indexing performance., Comment: Accepted by Tsinghua Science and Technology

As one of the most useful online processing techniques, the theta-join operation has been utilized by many applications to fully excavate the relationships between data streams in various scenarios. As such, constant research efforts have been put to optimize its performance in the distributed environment, which is typically characterized by reducing the number of Cartesian products as much as possible. In this article, we design and implement a novel fast theta-join algorithm, called Prefap, by developing two distinct techniques - prefiltering and amalgamated partitioning-based on the state-of-the-art FastThetaJoin algorithm to optimize the efficiency of the theta-join operation. Firstly, we develop a prefiltering strategy before data streams are partitioned to reduce the amount of data to be involved and benefit a more fine-grained partitioning. Secondly, to avoid the data streams being partitioned in a coarse-grained isolated manner and improve the quality of the partition-level filtering, we introduce an amalgamated partitioning mechanism that can amalgamate the partitioning boundaries of two data streams to assist a fine-grained partitioning. With the integration of these two techniques into the existing FastThetaJoin algorithm, we design and implement a new framework to achieve a decreased number of Cartesian products and a higher theta-join efficiency. By comparing with existing algorithms, FastThetaJoin in particular, we evaluate the performance of Prefap on both synthetic and real data streams from two-way to multiway theta-join to demonstrate its superiority., Comment: Accepted by Concurrency and Computation: Practice and Experience

Removing bias while keeping all task-relevant information is challenging for fair representation learning methods since they would yield random or degenerate representations w.r.t. labels when the sensitive attributes correlate with labels. Existing works proposed to inject the label information into the learning procedure to overcome such issues. However, the assumption that the observed labels are clean is not always met. In fact, label bias is acknowledged as the primary source inducing discrimination. In other words, the fair pre-processing methods ignore the discrimination encoded in the labels either during the learning procedure or the evaluation stage. This contradiction puts a question mark on the fairness of the learned representations. To circumvent this issue, we explore the following question: \emph{Can we learn fair representations predictable to latent ideal fair labels given only access to unreliable labels?} In this work, we propose a \textbf{D}e-\textbf{B}iased \textbf{R}epresentation Learning for \textbf{F}airness (DBRF) framework which disentangles the sensitive information from non-sensitive attributes whilst keeping the learned representations predictable to ideal fair labels rather than observed biased ones. We formulate the de-biased learning framework through information-theoretic concepts such as mutual information and information bottleneck. The core concept is that DBRF advocates not to use unreliable labels for supervision when sensitive information benefits the prediction of unreliable labels. Experiment results over both synthetic and real-world data demonstrate that DBRF effectively learns de-biased representations towards ideal labels.

Few-shot class-incremental learning (FSCIL) has been proposed aiming to enable a deep learning system to incrementally learn new classes with limited data. Recently, a pioneer claims that the commonly used replay-based method in class-incremental learning (CIL) is ineffective and thus not preferred for FSCIL. This has, if truth, a significant influence on the fields of FSCIL. In this paper, we show through empirical results that adopting the data replay is surprisingly favorable. However, storing and replaying old data can lead to a privacy concern. To address this issue, we alternatively propose using data-free replay that can synthesize data by a generator without accessing real data. In observing the the effectiveness of uncertain data for knowledge distillation, we impose entropy regularization in the generator training to encourage more uncertain examples. Moreover, we propose to relabel the generated data with one-hot-like labels. This modification allows the network to learn by solely minimizing the cross-entropy loss, which mitigates the problem of balancing different objectives in the conventional knowledge distillation approach. Finally, we show extensive experimental results and analysis on CIFAR-100, miniImageNet and CUB-200 to demonstrate the effectiveness of our proposed one., Comment: Accepted by ECCV 2022