Recent years have seen a strong uptick in both the prevalence and real-world consequences of false information spread through online platforms. At the same time, encrypted messaging systems such as WhatsApp, Signal, and Telegram, are rapidly gaining
popularity as users seek increased privacy in their digital lives. The challenge we address is how to combat the viral spread of misinformation without compromising privacy. Our FACTS system tracks user complaints on messages obliviously, only revealing the messages contents and originator once sufficiently many complaints have been lodged. Our system is private, meaning it does not reveal anything about the senders or contents of messages which have received few or no complaints; secure, meaning there is no way for a malicious user to evade the system or gain an outsized impact over the complaint system; and scalable, as we demonstrate excellent practical efficiency for up to millions of complaints per day. Our main technical contribution is a new collaborative counting Bloom filter, a simple construction with difficult probabilistic analysis, which may have independent interest as a privacy-preserving randomized count sketch data structure. Compared to prior work on message flagging and tracing in end-to-end encrypted messaging, our novel contribution is the addition of a high threshold of multiple complaints that are needed before a message is audited or flagged. We present and carefully analyze the probabilistic performance of our data structure, provide a precise security definition and proof, and then measure the accuracy and scalability of our scheme via experimentation.
Fine-grained action recognition is attracting increasing attention due to the emerging demand of specific action understanding in real-world applications, whereas the data of rare fine-grained categories is very limited. Therefore, we propose the few
-shot fine-grained action recognition problem, aiming to recognize novel fine-grained actions with only few samples given for each class. Although progress has been made in coarse-grained actions, existing few-shot recognition methods encounter two issues handling fine-grained actions: the inability to capture subtle action details and the inadequacy in learning from data with low inter-class variance. To tackle the first issue, a human vision inspired bidirectional attention module (BAM) is proposed. Combining top-down task-driven signals with bottom-up salient stimuli, BAM captures subtle action details by accurately highlighting informative spatio-temporal regions. To address the second issue, we introduce contrastive meta-learning (CML). Compared with the widely adopted ProtoNet-based method, CML generates more discriminative video representations for low inter-class variance data, since it makes full use of potential contrastive pairs in each training episode. Furthermore, to fairly compare different models, we establish specific benchmark protocols on two large-scale fine-grained action recognition datasets. Extensive experiments show that our method consistently achieves state-of-the-art performance across evaluated tasks.
We introduce the task of open-vocabulary visual instance search (OVIS). Given an arbitrary textual search query, Open-vocabulary Visual Instance Search (OVIS) aims to return a ranked list of visual instances, i.e., image patches, that satisfies the s
earch intent from an image database. The term open vocabulary means that there are neither restrictions to the visual instance to be searched nor restrictions to the word that can be used to compose the textual search query. We propose to address such a search challenge via visual-semantic aligned representation learning (ViSA). ViSA leverages massive image-caption pairs as weak image-level (not instance-level) supervision to learn a rich cross-modal semantic space where the representations of visual instances (not images) and those of textual queries are aligned, thus allowing us to measure the similarities between any visual instance and an arbitrary textual query. To evaluate the performance of ViSA, we build two datasets named OVIS40 and OVIS1600 and also introduce a pipeline for error analysis. Through extensive experiments on the two datasets, we demonstrate ViSAs ability to search for visual instances in images not available during training given a wide range of textual queries including those composed of uncommon words. Experimental results show that ViSA achieves an mAP@50 of 21.9% on OVIS40 under the most challenging setting and achieves an mAP@6 of 14.9% on OVIS1600 dataset.
On-demand delivery has become increasingly popular around the world. Brick-and-mortar grocery stores, restaurants, and pharmacies are providing fast delivery services to satisfy the growing home delivery demand. Motivated by a large meal and grocery
delivery company, we model and solve a multiperiod driver dispatching and routing problem for last-mile delivery systems where on-time performance is the main target. The operator of this system needs to dispatch a set of drivers and specify their delivery routes in a stochastic environment, in which random demand arrives over a fixed number of periods. The resulting dynamic program is challenging to solve due to the curse of dimensionality. We propose a novel approximation framework to approximate the value function via a simplified dispatching program. We then develop efficient exact algorithms for this problem based on Benders decomposition and column generation. We validate the superior performance of our framework and algorithms via extensive numerical experiments. Tested on a real-world data set, we quantify the value of adaptive dispatching and routing in on-time delivery and highlight the need of coordinating these two decisions in a dynamic setting. We show that dispatching multiple vehicles with short trips is preferable for on-time delivery, as opposed to sending a few vehicles with long travel times.
This paper describes our approach to DSTC 9 Track 2: Cross-lingual Multi-domain Dialog State Tracking, the task goal is to build a Cross-lingual dialog state tracker with a training set in rich resource language and a testing set in low resource lang
uage. We formulate a method for joint learning of slot operation classification task and state tracking task respectively. Furthermore, we design a novel mask mechanism for fusing contextual information about dialogue, the results show the proposed model achieves excellent performance on DSTC Challenge II with a joint accuracy of 62.37% and 23.96% in MultiWOZ(en - zh) dataset and CrossWOZ(zh - en) dataset, respectively.
Low-rankness is important in the hyperspectral image (HSI) denoising tasks. The tensor nuclear norm (TNN), defined based on the tensor singular value decomposition, is a state-of-the-art method to describe the low-rankness of HSI. However, TNN ignore
s some of the physical meanings of HSI in tackling the denoising tasks, leading to suboptimal denoising performance. In this paper, we propose the multi-modal and frequency-weighted tensor nuclear norm (MFWTNN) and the non-convex MFWTNN for HSI denoising tasks. Firstly, we investigate the physical meaning of frequency components and reconsider their weights to improve the low-rank representation ability of TNN. Meanwhile, we also consider the correlation among two spatial dimensions and the spectral dimension of HSI and combine the above improvements to TNN to propose MFWTNN. Secondly, we use non-convex functions to approximate the rank function of the frequency tensor and propose the NonMFWTNN to relax the MFWTNN better. Besides, we adaptively choose bigger weights for slices mainly containing noise information and smaller weights for slices containing profile information. Finally, we develop the efficient alternating direction method of multiplier (ADMM) based algorithm to solve the proposed models, and the effectiveness of our models are substantiated in simulated and real HSI datasets.
This paper mainly uses the nonnegative continuous function ${zeta_n(r)}_{n=0}^{infty}$ to redefine the Bohr radius for the class of analytic functions satisfying $real f(z)<1$ in the unit disk $|z|<1$ and redefine the Bohr radius of the alternating s
eries $A_f(r)$ with analytic functions $f$ of the form $f(z)=sum_{n=0}^{infty}a_{pn+m}z^{pn+m}$ in $|z|<1$. In the latter case, one can also get information about Bohr radius for even and odd analytic functions. Moreover, the relationships between the majorant series $M_f(r)$ and the odd and even bits of $f(z)$ are also established. We will prove that most of results are sharp.
Multiple signal classification (MUSIC) has been widely applied in multiple-input multiple-output (MIMO) receivers for direction-of-arrival (DOA) estimation. To reduce the cost of radio frequency (RF) chains operating at millimeter-wave bands, hybrid
analog-digital structure has been adopted in massive MIMO transceivers. In this situation, the received signals at the antennas are unavailable to the digital receiver, and as a consequence, the spatial covariance matrix (SCM), which is essential in MUSIC algorithm, cannot be obtained using traditional sample average approach. Based on our previous work, we propose a novel algorithm for SCM reconstruction in hybrid massive MIMO systems with multiple RF chains. By switching the analog beamformers to a group of predetermined DOAs, SCM can be reconstructed through the solutions of a set of linear equations. In addition, based on insightful analysis on that linear equations, a low-complexity algorithm, as well as a careful selection of the predetermined DOAs, will be also presented in this paper. Simulation results show that the proposed algorithms can reconstruct the SCM accurately so that MUSIC algorithm can be well used for DOA estimation in hybrid massive MIMO systems with multiple RF chains.
Bi-level optimization model is able to capture a wide range of complex learning tasks with practical interest. Due to the witnessed efficiency in solving bi-level programs, gradient-based methods have gained popularity in the machine learning communi
ty. In this work, we propose a new gradient-based solution scheme, namely, the Bi-level Value-Function-based Interior-point Method (BVFIM). Following the main idea of the log-barrier interior-point scheme, we penalize the regularized value function of the lower level problem into the upper level objective. By further solving a sequence of differentiable unconstrained approximation problems, we consequently derive a sequential programming scheme. The numerical advantage of our scheme relies on the fact that, when gradient methods are applied to solve the approximation problem, we successfully avoid computing any expensive Hessian-vector or Jacobian-vector product. We prove the convergence without requiring any convexity assumption on either the upper level or the lower level objective. Experiments demonstrate the efficiency of the proposed BVFIM on non-convex bi-level problems.
The skyrmion, which is characterised by a topological integer, is a structure that is topologically stable against local disturbances. The huge potential of skyrmions for use in magnetic storage systems has drawn considerable research interest among
physicists. Recently, the optical skyrmion was discovered and has some excellent properties. However, these optical skyrmions have been observed, for example, in surface plasmons that consist of evanescent waves. This type of optical skyrmion is difficult to manipulate and also difficult to apply in practice. In this work, we realise several skyrmionic optical structures with different skyrmion numbers in a free-space linear optical system. Because of the convenience of operation using free-space optics, with the exception of the original applications of skyrmions, skyrmionic optical structures can also be applied widely, e.g. to enable manipulation of tiny objects or propagation over long distances.
