Search

Quantum computing has emerged as a transformative tool for future data management. Classical problems in database domains, including query optimization, data integration, and transaction management, have recently been addressed using quantum computing techniques. This tutorial aims to establish the theoretical foundation essential for enhancing methodologies and practical implementations in this line of research. Moreover, this tutorial takes a forward-looking approach by delving into recent strides in quantum internet technologies and the nonlocality theory. We aim to shed light on the uncharted territory of future data systems tailored for the quantum internet., Comment: Accepted at the 40th IEEE International Conference on Data Engineering (ICDE 2024)

The report demonstrates the benefits (in terms of improved claims loss modeling) of harnessing the value of Federated Learning (FL) to learn a single model across multiple insurance industry datasets without requiring the datasets themselves to be shared from one company to another. The application of FL addresses two of the most pressing concerns: limited data volume and data variety, which are caused by privacy concerns, the rarity of claim events, the lack of informative rating factors, etc.. During each round of FL, collaborators compute improvements on the model using their local private data, and these insights are combined to update a global model. Such aggregation of insights allows for an increase to the effectiveness in forecasting claims losses compared to models individually trained at each collaborator. Critically, this approach enables machine learning collaboration without the need for raw data to leave the compute infrastructure of each respective data owner. Additionally, the open-source framework, OpenFL, that is used in our experiments is designed so that it can be run using confidential computing as well as with additional algorithmic protections against leakage of information via the shared model updates. In such a way, FL is implemented as a privacy-enhancing collaborative learning technique that addresses the challenges posed by the sensitivity and privacy of data in traditional machine learning solutions. This paper's application of FL can also be expanded to other areas including fraud detection, catastrophe modeling, etc., that have a similar need to incorporate data privacy into machine learning collaborations. Our framework and empirical results provide a foundation for future collaborations among insurers, regulators, academic researchers, and InsurTech experts.

While progress has been made in the domain of video-language understanding, current state-of-the-art algorithms are still limited in their ability to understand videos at high levels of abstraction, such as news-oriented videos. Alternatively, humans easily amalgamate information from video and language to infer information beyond what is visually observable in the pixels. An example of this is watching a news story, where the context of the event can play as big of a role in understanding the story as the event itself. Towards a solution for designing this ability in algorithms, we present a large-scale analysis on an in-house dataset collected by the Reuters News Agency, called Reuters Video-Language News (ReutersViLNews) dataset which focuses on high-level video-language understanding with an emphasis on long-form news. The ReutersViLNews Dataset consists of long-form news videos collected and labeled by news industry professionals over several years and contains prominent news reporting from around the world. Each video involves a single story and contains action shots of the actual event, interviews with people associated with the event, footage from nearby areas, and more. ReutersViLNews dataset contains videos from seven subject categories: disaster, finance, entertainment, health, politics, sports, and miscellaneous with annotations from high-level to low-level, title caption, visual video description, high-level story description, keywords, and location. We first present an analysis of the dataset statistics of ReutersViLNews compared to previous datasets. Then we benchmark state-of-the-art approaches for four different video-language tasks. The results suggest that news-oriented videos are a substantial challenge for current video-language understanding algorithms and we conclude by providing future directions in designing approaches to solve the ReutersViLNews dataset.

This work investigates the oracle separation between the physically motivated complexity class of noisy quantum circuits, inspired by definitions such as those presented by Chen, Cotler, Huang, and Li (2022). We establish that with a constant error rate, separation can be achieved in terms of NP. When the error rate is $\Omega(\log n/n)$, we can extend this result to the separation of PH. Notably, our oracles, in all separations, do not necessitate error correction schemes or fault tolerance, as all quantum circuits are of constant depth. This indicates that even quantum computers with minor errors, without error correction, may surpass classical complexity classes under various scenarios and assumptions. We also explore various common noise settings and present new classical hardness results, generalizing those found in studies by Raz and Tal (2022) and Bassirian, Bouland, Fefferman, Gunn, and Tal (2021), which are of independent interest.

Tyron Tai-Chun Chou,1,* Hsin-Chien Hsu,1– 3,* Chih-Wen Twu,4,5 Wen-Kuan Huang,6 Hung-Meng Huang,1,7 Shih-Han Weng,8 Ming-Chih Chen2 1Department of Otolaryngology, Taipei City Hospital, Taipei City, Taiwan; 2Graduate Institute of Business Administration, College of Management, Fu Jen Catholic University, New Taipei City, Taiwan; 3General Education Center, University of Taipei, Taipei City, Taiwan; 4Department of Otorhinolaryngology, Head and Neck Surgery, Changhua Christian Hospital, Changhua County, Taiwan; 5Department of Post-Baccalaureate Medicine, National Chung Hsing University, Taichung City, Taiwan; 6Division of Hematology/Oncology, Department of Internal Medicine, Chang Gung Memorial Hospital at Linkou, Chang Gung University College of Medicine, Taoyuan City, Taiwan; 7Department of Otolaryngology, School of Medicine, College of Medicine, Taipei Medical University, Taipei City, Taiwan; 8Department of Education and Research, Taipei City Hospital, Taipei City, Taiwan*These authors contributed equally to this workCorrespondence: Ming-Chih Chen, Institute of Business Administration, College of Management, Fu Jen Catholic University, No. 510, Zhongzheng Road, Xinzhuang District, New Taipei City, 242062, Taiwan, Email 081438@mail.fju.edu.twBackground: Obstructive sleep apnea syndrome (OSAS) is a common disorder associated with serious sequelae. The current gold standard diagnostic method, polysomnography, is costly and time consuming and requires patients to stay overnight at a facility.Aim: This study aimed to reveal the prevalence of OSAS in general adult population using a home sleep test (HST) during the coronavirus disease 2019 (COVID-19) pandemic.Methods: This prospective cohort study was conducted by the Department of Otolaryngology, Taipei City Hospital, Taipei, Taiwan, between January 2020 and December 2021. A total of 1372 patients aged 30– 70 years completed an HST using a Type 3 portable sleep monitor (PM). The apnea-hypopnea

Existing dense or paragraph video captioning approaches rely on holistic representations of videos, possibly coupled with learned object/action representations, to condition hierarchical language decoders. However, they fundamentally lack the commonsense knowledge of the world required to reason about progression of events, causality, and even the function of certain objects within a scene. To address this limitation we propose a novel video captioning Transformer-based model, that takes into account both implicit (visuo-lingual and purely linguistic) and explicit (knowledge-base) commonsense knowledge. We show that these forms of knowledge, in isolation and in combination, enhance the quality of produced captions. Further, inspired by imitation learning, we propose a new task of instruction generation, where the goal is to produce a set of linguistic instructions from a video demonstration of its performance. We formalize the task using the ALFRED dataset [54] generated using an AI2-THOR environment. While instruction generation is conceptually similar to paragraph captioning, it differs in the fact that it exhibits stronger object persistence, as well as spatially-aware and causal sentence structure. We show that our commonsense knowledge enhanced approach produces significant improvements on this task (up to 57% in METEOR and 8.5% in CIDEr), as well as the state-of-the-art result on more traditional video captioning in the ActivityNet Captions dataset [29]., Comment: The paper is under consideration at Computer Vision and Image Understanding Journal

We propose an application for near-term quantum devices: namely, generating cryptographically certified random bits, to use (for example) in proof-of-stake cryptocurrencies. Our protocol repurposes the existing "quantum supremacy" experiments, based on random circuit sampling, that Google and USTC have successfully carried out starting in 2019. We show that, whenever the outputs of these experiments pass the now-standard Linear Cross-Entropy Benchmark (LXEB), under plausible hardness assumptions they necessarily contain $\Omega(n)$ min-entropy, where $n$ is the number of qubits. To achieve a net gain in randomness, we use a small random seed to produce pseudorandom challenge circuits. In response to the challenge circuits, the quantum computer generates output strings that, after verification, can then be fed into a randomness extractor to produce certified nearly-uniform bits -- thereby "bootstrapping" from pseudorandomness to genuine randomness. We prove our protocol sound in two senses: (i) under a hardness assumption called Long List Quantum Supremacy Verification, which we justify in the random oracle model, and (ii) unconditionally in the random oracle model against an eavesdropper who could share arbitrary entanglement with the device. (Note that our protocol's output is unpredictable even to a computationally unbounded adversary who can see the random oracle.) Currently, the central drawback of our protocol is the exponential cost of verification, which in practice will limit its implementation to at most $n\sim 60$ qubits, a regime where attacks are expensive but not impossible. Modulo that drawback, our protocol appears to be the only practical application of quantum computing that both requires a QC and is physically realizable today., Comment: 84 pages

This paper explains the generalization power of a deep neural network (DNN) from the perspective of interactive concepts. Many recent studies have quantified a clear emergence of interactive concepts encoded by the DNN, which have been observed on different DNNs during the learning process. Therefore, in this paper, we investigate the generalization power of each interactive concept, and we use the generalization power of different interactive concepts to explain the generalization power of the entire DNN. Specifically, we define the complexity of each interactive concept. We find that simple concepts can be better generalized to testing data than complex concepts. The DNN with strong generalization power usually learns simple concepts more quickly and encodes fewer complex concepts. More crucially, we discover the detouring dynamics of learning complex concepts, which explain both the high learning difficulty and the low generalization power of complex concepts.

In this work, we study the computational complexity of quantum determinants, a $q$-deformation of matrix permanents: Given a complex number $q$ on the unit circle in the complex plane and an $n\times n$ matrix $X$, the $q$-permanent of $X$ is defined as $$\mathrm{Per}_q(X) = \sum_{\sigma\in S_n} q^{\ell(\sigma)}X_{1,\sigma(1)}\ldots X_{n,\sigma(n)},$$ where $\ell(\sigma)$ is the inversion number of permutation $\sigma$ in the symmetric group $S_n$ on $n$ elements. The function family generalizes determinant and permanent, which correspond to the cases $q=-1$ and $q=1$ respectively. For worst-case hardness, by Liouville's approximation theorem and facts from algebraic number theory, we show that for primitive $m$-th root of unity $q$ for odd prime power $m=p^k$, exactly computing $q$-permanent is $\mathsf{Mod}_p\mathsf{P}$-hard. This implies that an efficient algorithm for computing $q$-permanent results in a collapse of the polynomial hierarchy. Next, we show that computing $q$-permanent can be achieved using an oracle that approximates to within a polynomial multiplicative error and a membership oracle for a finite set of algebraic integers. From this, an efficient approximation algorithm would also imply a collapse of the polynomial hierarchy. By random self-reducibility, computing $q$-permanent remains to be hard for a wide range of distributions satisfying a property called the strong autocorrelation property. Specifically, this is proved via a reduction from $1$-permanent to $q$-permanent for $O(1/n^2)$ points $z$ on the unit circle. Since the family of permanent functions shares common algebraic structure, various techniques developed for the hardness of permanent can be generalized to $q$-permanents.

We study quantum algorithms that learn properties of a matrix using queries that return its action on an input vector. We show that for various problems, including computing the trace, determinant, or rank of a matrix or solving a linear system that it specifies, quantum computers do not provide an asymptotic speedup over classical computation. On the other hand, we show that for some problems, such as computing the parities of rows or columns or deciding if there are two identical rows or columns, quantum computers provide exponential speedup. We demonstrate this by showing equivalence between models that provide matrix-vector products, vector-matrix products, and vector-matrix-vector products, whereas the power of these models can vary significantly for classical computation.

As quantum computing progresses steadily from theory into practice, programmers will face a common problem: How can they be sure that their code does what they intend it to do? This paper presents encouraging results in the application of mechanized proof to the domain of quantum programming in the context of the SQIR development. It verifies the correctness of a range of a quantum algorithms including Grover’s algorithm and quantum phase estimation, a key component of Shor’s algorithm. In doing so, it aims to highlight both the successes and challenges of formal verification in the quantum context and motivate the theorem proving community to target quantum computing as an application domain.

Prodrug activator gene therapy mediated by murine leukemia virus (MLV)-based retroviral replicating vectors (RRV) was previously shown to be highly effective in killing glioma cells both in culture and in vivo. To avoid receptor interference and enable dual vector co-infection with MLV-RRV, we have developed another RRV based on gibbon ape leukemia virus (GALV) that also shows robust replicative spread in a wide variety of tumor cells. We evaluated the potential of GALV-based RRV as a cancer therapeutic agent by incorporating yeast cytosine deaminase (CD) and E. coli nitroreductase (NTR) prodrug activator genes into the vector. The expression of CD and NTR genes from GALV-RRV achieved highly efficient delivery of these prodrug activator genes to RG-2 glioma cells, resulting in enhanced cytotoxicity after administering their respective prodrugs 5-fluorocytosine and CB1954 in vitro. In an immune-competent intracerebral RG-2 glioma model, GALV-mediated CD and NTR gene therapy both significantly suppressed tumor growth with CB1954 administration after a single injection of vector supernatant. However, NTR showed greater potency than CD, with control animals receiving GALV-NTR vector alone (i.e., without CB1954 prodrug) showing extensive tumor growth with a median survival time of 17.5 days, while animals receiving GALV-NTR and CB1954 showed significantly prolonged survival with a median survival time of 30 days. In conclusion, GALV-RRV enabled high-efficiency gene transfer and persistent expression of NTR, resulting in efficient cell killing, suppression of tumor growth, and prolonged survival upon CB1954 administration. This validates the use of therapeutic strategies employing this prodrug activator gene to arm GALV-RRV, and opens the door to the possibility of future combination gene therapy with CD-armed MLV-RRV, as the latter vector is currently being evaluated in clinical trials.

Although machine learning (ML) has shown promise across disciplines, out-of-sample generalizability is concerning. This is currently addressed by sharing multi-site data, but such centralization is challenging/infeasible to scale due to various limitations. Federated ML (FL) provides an alternative paradigm for accurate and generalizable ML, by only sharing numerical model updates. Here we present the largest FL study to-date, involving data from 71 sites across 6 continents, to generate an automatic tumor boundary detector for the rare disease of glioblastoma, reporting the largest such dataset in the literature (n = 6, 314). We demonstrate a 33% delineation improvement for the surgically targetable tumor, and 23% for the complete tumor extent, over a publicly trained model. We anticipate our study to: 1) enable more healthcare studies informed by large diverse data, ensuring meaningful results for rare diseases and underrepresented populations, 2) facilitate further analyses for glioblastoma by releasing our consensus model, and 3) demonstrate the FL effectiveness at such scale and task-complexity as a paradigm shift for multi-site collaborations, alleviating the need for data-sharing.

An engaging and provocative question can open up a great conversation. In this work, we explore a novel scenario: a conversation agent views a set of the user's photos (for example, from social media platforms) and asks an engaging question to initiate a conversation with the user. The existing vision-to-question models mostly generate tedious and obvious questions, which might not be ideals conversation starters. This paper introduces a two-phase framework that first generates a visual story for the photo set and then uses the story to produce an interesting question. The human evaluation shows that our framework generates more response-provoking questions for starting conversations than other vision-to-question baselines., Comment: Accepted as a full talk paper on AAAI-DEEPDIAL'21

We present two protocols for classical verification of quantum depth. Our protocols allow a purely classical verifier to distinguish devices with different quantum circuit depths even in the presence of classical computation. We show that a device with quantum circuit depth at most d will be rejected by the verifier even if the prover applies additional polynomial-time classical computation to cheat. On the other hand, the verifier accepts a device which has quantum circuit depth d' for some d'>d. In our first protocol, we introduce an additional untrusted quantum machine which shares entanglements with the target machine. Applying a robust self-test, our first protocol certifies the depth of the target machine with information theoretic security and nearly optimal separation. The protocol relies on the oracle separation problem for quantum depth by Chia, Chung and Lai [STOC 2020] and a transformation from an oracle separation problem to a two-player non-local game. Our second protocol certifies the quantum depth of a single device based on quantum hardness of learning with errors. The protocol relies on the noisy trapdoor claw-free function family and the idea of pointer chasing to force the prover to keep quantum coherence until all preceding message exchanges are completed. To our knowledge, we give the first constructions for distinguishing hybrid quantum-classical computers with different circuit depths in unrelativized models., Comment: 39 pages

This paper aims to theoretically analyze the complexity of feature transformations encoded in piecewise linear DNNs with ReLU layers. We propose metrics to measure three types of complexities of transformations based on the information theory. We further discover and prove the strong correlation between the complexity and the disentanglement of transformations. Based on the proposed metrics, we analyze two typical phenomena of the change of the transformation complexity during the training process, and explore the ceiling of a DNN's complexity. The proposed metrics can also be used as a loss to learn a DNN with the minimum complexity, which also controls the over-fitting level of the DNN and influences adversarial robustness, adversarial transferability, and knowledge consistency. Comprehensive comparative studies have provided new perspectives to understand the DNN.

Deep learning models have been deployed in numerous real-world applications such as autonomous driving and surveillance. However, these models are vulnerable in adversarial environments. Backdoor attack is emerging as a severe security threat which injects a backdoor trigger into a small portion of training data such that the trained model behaves normally on benign inputs but gives incorrect predictions when the specific trigger appears. While most research in backdoor attacks focuses on image classification, backdoor attacks on object detection have not been explored but are of equal importance. Object detection has been adopted as an important module in various security-sensitive applications such as autonomous driving. Therefore, backdoor attacks on object detection could pose severe threats to human lives and properties. We propose four kinds of backdoor attacks for object detection task: 1) Object Generation Attack: a trigger can falsely generate an object of the target class; 2) Regional Misclassification Attack: a trigger can change the prediction of a surrounding object to the target class; 3) Global Misclassification Attack: a single trigger can change the predictions of all objects in an image to the target class; and 4) Object Disappearance Attack: a trigger can make the detector fail to detect the object of the target class. We develop appropriate metrics to evaluate the four backdoor attacks on object detection. We perform experiments using two typical object detection models -- Faster-RCNN and YOLOv3 on different datasets. More crucially, we demonstrate that even fine-tuning on another benign dataset cannot remove the backdoor hidden in the object detection model. To defend against these backdoor attacks, we propose Detector Cleanse, an entropy-based run-time detection framework to identify poisoned testing samples for any deployed object detector.

An astigmatic optical profilometer is a precision instrument with advantages such as high resolution, high bandwidth, a compact size, and low cost. However, current astigmatic optical profilometers measure only surface morphology, and their potential for capturing subsurface information remains underutilized. In this study, we developed a method for measuring the thickness of transparent thin films with an astigmatic optical profilometer. Experimental results demonstrate that the thickness of transparent films tens of micrometers thick can be accurately measured. The maximum thickness measurable through our system is approximately 100 µm, which may be increased to 1.2 mm through the use of a scanner with a greater travel range. A coupling problem occurs for films <25 µm in thickness. However, to solve this problem, we devised a decoupling method, which was experimentally implemented to successfully measure a 18-µm-thick film. Moreover, the ability to obtain 3D images, including of both the upper and lower surfaces, was demonstrated.

We study quantum algorithms that learn properties of a matrix using queries that return its action on an input vector. We show that for various problems, including computing the trace, determinant, or rank of a matrix or solving a linear system that it specifies, quantum computers do not provide an asymptotic speedup over classical computation. On the other hand, we show that for some problems, such as computing the parities of rows or columns or deciding if there are two identical rows or columns, quantum computers provide exponential speedup. We demonstrate this by showing equivalence between models that provide matrix-vector products, vector-matrix products, and vector-matrix-vector products, whereas the power of these models can vary significantly for classical computation.

Sepsis develops from a serious microbial infection that causes the immune system to go into overdrive. The major microorganisms that induce sepsis are Gram-negative bacteria with lipopolysaccharide (LPS) in their cell walls. Nitric oxide (NO) and cyclooxygenase-2 (COX-2) are the key factors involved in the LPS-induced pro-inflammatory process. This study aimed to evaluate the effects of polyphenol Tellimagrandin II (TGII) on anti-inflammatory activity and its underlying basic mechanism in murine macrophage cell line RAW 264.7 and human monocyte-derived macrophages. Macrophages with more than 90% cell viability were found in the cytotoxicity assay under 50 mu M TGII. Pre- or post-treatment with TGII significantly reduced LPS-induced inducible nitric oxide synthase (NOS2) protein and mRNA expression, reducing LPS-induced COX-2 protein. Downstream of NOS2 and COX-2, NO and prostaglandin E2 (PGE2) were significantly inhibited by TGII. Upstream of NOS2 and COX-2, phospho-p65, c-fos and phospho-c-jun were also reduced after pre-treatment with TGII. Mitogen-activated protein kinases (MAPKs) are also critical to nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kappa B) stimulation, and phospho-p38 expression was found to have been blocked by TGII. TGII efficiently reduces LPS-induced NO production and its upstream regulatory factors, suggesting that TGII may be a potential therapeutic agent for sepsis and other inflammatory diseases.

Despite recent advances of data acquisition and algorithms development, machine learning (ML) faces tremendous challenges to being adopted in practical catalyst design, largely due to its limited generalizability and poor explainability. Herein, we develop a theory-infused neural network (TinNet) approach that integrates deep learning algorithms with the wellestablished d-band theory of chemisorption for reactivity prediction of transition-metal surfaces. With simple adsorbates (e.g., *OH, *O, and *N) at active site ensembles as representative descriptor species, we demonstrate that the TinNet is on par with purely data-driven ML methods in prediction performance while being inherently interpretable. Incorporation of scientific knowledge of physical interactions into learning from data sheds further light on the nature of chemical bonding and opens up new avenues for ML discovery of novel motifs with desired catalytic properties.

Understanding the nature of chemical bonding and its variation in strength across physically tunable factors is important for the development of novel catalytic materials. One way to speed up this process is to employ machine learning (ML) algorithms with online data repositories curated from high-throughput experiments or quantum-chemical simulations. Despite the reasonable predictive performance of ML models for predicting reactivity properties of solid surfaces, the ever-growing complexity of modern algorithms, e.g., deep learning, makes them black boxes with little to no explanation. In this Perspective, we discuss recent advances of interpretable ML for opening up these black boxes from the standpoints of feature engineering, algorithm development, and post hoc analysis. We underline the pivotal role of interpretability as the foundation of next-generation ML algorithms and emerging AI platforms for driving discoveries across scientific disciplines.

Quantum computing devices can solve problems that are infeasible for classical computers. While rigorously proving speedups over existing classical algorithms demonstrates the usefulness of quantum computers, analyzing the limits on efficient processes for computational tasks allows us to better understand the power of quantum computation. Indeed, hard problems for quantum computers also enable useful cryptographic applications. In this dissertation, we aim to understand the limits on efficient quantum computation and base applications on hard problems for quantum computers. We consider models in which a classical machine can leverage the power of a quantum device, which may be affected by noise or behave adversarially. We present protocols and tools for detecting errors in a quantum machine and estimate how serious the deviation is. We construct a non-interactive protocol that enables a purely classical party to delegate any quantum computation to an untrusted quantum prover. In the setting of error-prone quantum hardware, we employ formal methods to construct a logical system for reasoning about the robustness of a quantum algorithm design. We also study the limits of ideal quantum computers for computational tasks and give asymptotically optimal algorithms. In particular, we give quantum algorithms which provide speedups for the polynomial interpolation problem and show their optimality. Finally, we study the performance of quantum algorithms that learn properties of a matrix using queries that return its action on an input vector. In particular, we show that for various linear algebra problems, there is no quantum speedup, while for some problems, exponential speedups can be achieved.

An astigmatic optical profilometer provides numerous advantages such as compact size, low cost, and high resolution. In 2018, a z-axis modulation mode of the astigmatic optical profilometer was developed to realize quantitative height measurement for a surface comprising complex materials. However, the current imaging rate of the z-axis modulation mode is time consuming. In this work, a resonant scanner was proposed to execute a high-speed z-axis scanning motion. By inducing the resonant mode, the resonant scanner enabled both a large travel range of over 87 m and high oscillation frequency of 1.576 kHz. Additionally, a data analysis process was proposed to calibrate the nonlinear movement of the resonant scanner. Experimental results demonstrated that the developed optical profilometer successfully captured the quantitative height and reflectivity of the complex material comprising chrome patterns on a glass substrate. Furthermore, an imaging rate of 256 s/frame was achieved, which was approximately 50 times faster than that of a previous system employing a commercial flexure-guided scanner.

Federated learning (FL) is a computational paradigm that enables organizations to collaborate on machine learning (ML) projects without sharing sensitive data, such as, patient records, financial data, or classified secrets. Open Federated Learning (OpenFL https://github.com/intel/openfl) is an open-source framework for training ML algorithms using the data-private collaborative learning paradigm of FL. OpenFL works with training pipelines built with both TensorFlow and PyTorch, and can be easily extended to other ML and deep learning frameworks. Here, we summarize the motivation and development characteristics of OpenFL, with the intention of facilitating its application to existing ML model training in a production environment. Finally, we describe the first use of the OpenFL framework to train consensus ML models in a consortium of international healthcare organizations, as well as how it facilitates the first computational competition on FL.

This research takes a cross-disciplinary approach to examining the mind-modelling representations in picture books for children. Working with Theory of Mind research, the thesis develops a novel model and typology to systematically analyse the mind-reading processes of characters and their constructions in the sample texts. The in-depth analyses of the words and images in nineteen picture books from Australia, the United States, the United Kingdom, South Africa and Taiwan offer insight into understanding characters' emotional and psychological states. This thesis contributes new knowledge on mind-modelling in picture books and offers guidelines for further research into classroom applications.

Variational Quantum Circuits (VQCs), or the so-called quantum neural-networks, are predicted to be one of the most important near-term quantum applications, not only because of their similar promises as classical neural-networks, but also because of their feasibility on near-term noisy intermediate-size quantum (NISQ) machines. The need for gradient information in the training procedure of VQC applications has stimulated the development of auto-differentiation techniques for quantum circuits. We propose the first formalization of this technique, not only in the context of quantum circuits but also for imperative quantum programs (e.g., with controls), inspired by the success of differentiable programming languages in classical machine learning. In particular, we overcome a few unique difficulties caused by exotic quantum features (such as quantum no-cloning) and provide a rigorous formulation of differentiation applied to bounded-loop imperative quantum programs, its code-transformation rules, as well as a sound logic to reason about their correctness. Moreover, we have implemented our code transformation in OCaml and demonstrated the resource-efficiency of our scheme both analytically and empirically. We also conduct a case study of training a VQC instance with controls, which shows the advantage of our scheme over existing auto-differentiation for quantum circuits without controls., Comment: Codes are available at https://github.com/LibertasSpZ/adcompile

In this work, we introduce VQA 360, a novel task of visual question answering on 360 images. Unlike a normal field-of-view image, a 360 image captures the entire visual content around the optical center of a camera, demanding more sophisticated spatial understanding and reasoning. To address this problem, we collect the first VQA 360 dataset, containing around 17,000 real-world image-question-answer triplets for a variety of question types. We then study two different VQA models on VQA 360, including one conventional model that takes an equirectangular image (with intrinsic distortion) as input and one dedicated model that first projects a 360 image onto cubemaps and subsequently aggregates the information from multiple spatial resolutions. We demonstrate that the cubemap-based model with multi-level fusion and attention diffusion performs favorably against other variants and the equirectangular-based models. Nevertheless, the gap between the humans' and machines' performance reveals the need for more advanced VQA 360 algorithms. We, therefore, expect our dataset and studies to serve as the benchmark for future development in this challenging task. Dataset, code, and pre-trained models are available online., Comment: Accepted to WACV 2020

We consider the problem of Visual Question Answering (VQA). Given an image and a free-form, open-ended, question, expressed in natural language, the goal of VQA system is to provide accurate answer to this question with respect to the image. The task is challenging because it requires simultaneous and intricate understanding of both visual and textual information. Attention, which captures intra- and inter-modal dependencies, has emerged as perhaps the most widely used mechanism for addressing these challenges. In this paper, we propose an improved attention-based architecture to solve VQA. We incorporate an Attention on Attention (AoA) module within encoder-decoder framework, which is able to determine the relation between attention results and queries. Attention module generates weighted average for each query. On the other hand, AoA module first generates an information vector and an attention gate using attention results and current context; and then adds another attention to generate final attended information by multiplying the two. We also propose multimodal fusion module to combine both visual and textual information. The goal of this fusion module is to dynamically decide how much information should be considered from each modality. Extensive experiments on VQA-v2 benchmark dataset show that our method achieves the state-of-the-art performance., Comment: 8 pages

We investigate adversarial-sample generation methods from a frequency domain perspective and extend standard $l_{\infty}$ Projected Gradient Descent (PGD) to the frequency domain. The resulting method, which we call Spectral Projected Gradient Descent (SPGD), has better success rate compared to PGD during early steps of the method. Adversarially training models using SPGD achieves greater adversarial accuracy compared to PGD when holding the number of attack steps constant. The use of SPGD can, therefore, reduce the overhead of adversarial training when utilizing adversarial generation with a smaller number of steps. However, we also prove that SPGD is equivalent to a variant of the PGD ordinarily used for the $l_{\infty}$ threat model. This PGD variant omits the sign function which is ordinarily applied to the gradient. SPGD can, therefore, be performed without explicitly transforming into the frequency domain. Finally, we visualize the perturbations SPGD generates and find they use both high and low-frequency components, which suggests that removing either high-frequency components or low-frequency components is not an effective defense., Comment: 9 pages, 9 figures, SPAI'20, ACM ASIACCS 2020

As quantum computing progresses steadily from theory into practice, programmers will face a common problem: How can they be sure that their code does what they intend it to do? This paper presents encouraging results in the application of mechanized proof to the domain of quantum programming in the context of the SQIR development. It verifies the correctness of a range of a quantum algorithms including Grover's algorithm and quantum phase estimation, a key component of Shor's algorithm. In doing so, it aims to highlight both the successes and challenges of formal verification in the quantum context and motivate the theorem proving community to target quantum computing as an application domain., Comment: version 4 updated DOI (paper content is the same); version 3 (final version) has updated formatting and improved writing; version 2 includes updated acknowledgments and a new appendix with simple SQIR example programs

We present VOQC, the first fully verified optimizer for quantum circuits, written using the Coq proof assistant. Quantum circuits are expressed as programs in a simple, low-level language called SQIR, a simple quantum intermediate representation, which is deeply embedded in Coq. Optimizations and other transformations are expressed as Coq functions, which are proved correct with respect to a semantics of SQIR programs. SQIR uses a semantics of matrices of complex numbers, which is the standard for quantum computation, but treats matrices symbolically in order to reason about programs that use an arbitrary number of quantum bits. SQIR's careful design and our provided automation make it possible to write and verify a broad range of optimizations in VOQC, including full-circuit transformations from cutting-edge optimizers., Comment: This paper supercedes arXiv:1904.06319; version 2 includes additional results and improved formatting; version 3 is the final draft with additional formatting improvements and some restructuring

In a recent breakthrough, Mahadev constructed an interactive protocol that enables a purely classical party to delegate any quantum computation to an untrusted quantum prover. In this work, we show that this same task can in fact be performed non-interactively and in zero-knowledge. Our protocols result from a sequence of significant improvements to the original four-message protocol of Mahadev. We begin by making the first message instance-independent and moving it to an offline setup phase. We then establish a parallel repetition theorem for the resulting three-message protocol, with an asymptotically optimal rate. This, in turn, enables an application of the Fiat-Shamir heuristic, eliminating the second message and giving a non-interactive protocol. Finally, we employ classical non-interactive zero-knowledge (NIZK) arguments and classical fully homomorphic encryption (FHE) to give a zero-knowledge variant of this construction. This yields the first purely classical NIZK argument system for QMA, a quantum analogue of NP. We establish the security of our protocols under standard assumptions in quantum-secure cryptography. Specifically, our protocols are secure in the Quantum Random Oracle Model, under the assumption that Learning with Errors is quantumly hard. The NIZK construction also requires circuit-private FHE., Comment: 37 pages

While there are several widely used object detection datasets, current computer vision algorithms are still limited in conventional images. Such images narrow our vision in a restricted region. On the other hand, 360{\deg} images provide a thorough sight. In this paper, our goal is to provide a standard dataset to facilitate the vision and machine learning communities in 360{\deg} domain. To facilitate the research, we present a real-world 360{\deg} panoramic object detection dataset, 360-Indoor, which is a new benchmark for visual object detection and class recognition in 360{\deg} indoor images. It is achieved by gathering images of complex indoor scenes containing common objects and the intensive annotated bounding field-of-view. In addition, 360-Indoor has several distinct properties: (1) the largest category number (37 labels in total). (2) the most complete annotations on average (27 bounding boxes per image). The selected 37 objects are all common in indoor scene. With around 3k images and 90k labels in total, 360-Indoor achieves the largest dataset for detection in 360{\deg} images. In the end, extensive experiments on the state-of-the-art methods for both classification and detection are provided. We will release this dataset in the near future.

Estimating the volume of a convex body is a central problem in convex geometry and can be viewed as a continuous version of counting. We present a quantum algorithm that estimates the volume of an $n$-dimensional convex body within multiplicative error $\epsilon$ using $\tilde{O}(n^{3}+n^{2.5}/\epsilon)$ queries to a membership oracle and $\tilde{O}(n^{5}+n^{4.5}/\epsilon)$ additional arithmetic operations. For comparison, the best known classical algorithm uses $\tilde{O}(n^{4}+n^{3}/\epsilon^{2})$ queries and $\tilde{O}(n^{6}+n^{5}/\epsilon^{2})$ additional arithmetic operations. To the best of our knowledge, this is the first quantum speedup for volume estimation. Our algorithm is based on a refined framework for speeding up simulated annealing algorithms that might be of independent interest. This framework applies in the setting of "Chebyshev cooling", where the solution is expressed as a telescoping product of ratios, each having bounded variance. We develop several novel techniques when implementing our framework, including a theory of continuous-space quantum walks with rigorous bounds on discretization error. To complement our quantum algorithms, we also prove that volume estimation requires $\Omega(\sqrt n+1/\epsilon)$ quantum membership queries, which rules out the possibility of exponential quantum speedup in $n$ and shows optimality of our algorithm in $1/\epsilon$ up to poly-logarithmic factors., Comment: 61 pages, 8 figures. v2: Quantum query complexity improved to $\tilde{O}(n^{3}+n^{2.5}/\epsilon)$ and number of additional arithmetic operations improved to $\tilde{O}(n^{5}+n^{4.5}/\epsilon)$. v3: Improved Section 4.3.3 on nondestructive mean estimation and Section 6 on quantum lower bounds; various minor changes

We present sqire, a low-level language for quantum computing and verification. sqire uses a global register of quantum bits, allowing easy compilation to and from existing `quantum assembly' languages and simplifying the verification process. We demonstrate the power of sqire as an intermediate representation of quantum programs by verifying a number of useful optimizations, and we demonstrate sqire's use as a tool for general verification by proving several quantum programs correct., Comment: Superceded by arXiv:1912.02250

We consider the number of quantum queries required to determine the coefficients of a degree-d polynomial over F_q. A lower bound shown independently by Kane and Kutin and by Meyer and Pommersheim shows that d/2 + 1/2 quantum queries are needed to solve this problem with bounded error, whereas an algorithm of Boneh and Zhandry shows that d quantum queries are sufficient. We show that the lower bound is achievable: d/2 + 1/2 quantum queries suffice to determine the polynomial with bounded error. Furthermore, we show that d/2 + 1 queries suffice to achieve probability approaching 1 for large q. These upper bounds improve results of Boneh and Zhandry on the insecurity of cryptographic protocols against quantum attacks. We also show that our algorithm’s success probability as a function of the number of queries is precisely optimal. Furthermore, the algorithm can be implemented with gate complexity poly(log(q)) with negligible decrease in the success probability. We end with a conjecture about the quantum query complexity of multivariate polynomial interpolation.

Quantum computation is a topic of significant recent interest, with practical advances coming from both research and industry. A major challenge in quantum programming is dealing with errors (quantum noise) during execution. Because quantum resources (e.g., qubits) are scarce, classical error correction techniques applied at the level of the architecture are currently cost-prohibitive. But while this reality means that quantum programs are almost certain to have errors, there as yet exists no principled means to reason about erroneous behavior. This paper attempts to fill this gap by developing a semantics for erroneous quantum while-programs, as well as a logic for reasoning about them. This logic permits proving a property we have identified, called $\epsilon$-robustness, which characterizes possible "distance" between an ideal program and an erroneous one. We have proved the logic sound, and showed its utility on several case studies, notably: (1) analyzing the robustness of noisy versions of the quantum Bernoulli factory (QBF) and quantum walk (QW); (2) demonstrating the (in)effectiveness of different error correction schemes on single-qubit errors; and (3) analyzing the robustness of a fault-tolerant version of QBF., Comment: 34 pages, LaTeX; v2: fixed typos

For survival, a living agent must have the ability to assess risk (1) by temporally anticipating accidents before they occur, and (2) by spatially localizing risky regions in the environment to move away from threats. In this paper, we take an agent-centric approach to study the accident anticipation and risky region localization tasks. We propose a novel soft-attention Recurrent Neural Network (RNN) which explicitly models both spatial and appearance-wise non-linear interaction between the agent triggering the event and another agent or static-region involved. In order to test our proposed method, we introduce the Epic Fail (EF) dataset consisting of 3000 viral videos capturing various accidents. In the experiments, we evaluate the risk assessment accuracy both in the temporal domain (accident anticipation) and spatial domain (risky region localization) on our EF dataset and the Street Accident (SA) dataset. Our method consistently outperforms other baselines on both datasets.

Compared to current mobile networks, next-generation mobile networks are expected to support higher numbers of simultaneously connected devices and to achieve higher system spectrum efficiency and lower power consumption. To achieve these goals, we study the multi-sharing device-to-device (D2D) communication, which allows any cellular user equipment to share its radio resource with multiple D2D devices. We jointly consider resource block reuse and power control and then develop the MISS algorithm. Simulation results show that MISS performs very well in terms of transmission power consumption, system throughput, and the number of permitted D2D devices.

How many quantum queries are required to determine the coefficients of a degree-$d$ polynomial in $n$ variables? We present and analyze quantum algorithms for this multivariate polynomial interpolation problem over the fields $\mathbb{F}_q$, $\mathbb{R}$, and $\mathbb{C}$. We show that $k_{\mathbb{C}}$ and $2k_{\mathbb{C}}$ queries suffice to achieve probability $1$ for $\mathbb{C}$ and $\mathbb{R}$, respectively, where $k_{\mathbb{C}}=\smash{\lceil\frac{1}{n+1}{n+d\choose d}\rceil}$ except for $d=2$ and four other special cases. For $\mathbb{F}_q$, we show that $\smash{\lceil\frac{d}{n+d}{n+d\choose d}\rceil}$ queries suffice to achieve probability approaching $1$ for large field order $q$. The classical query complexity of this problem is $\smash{n+d\choose d}$, so our result provides a speedup by a factor of $n+1$, $\frac{n+1}{2}$, and $\frac{n+d}{d}$ for $\mathbb{C}$, $\mathbb{R}$, and $\mathbb{F}_q$, respectively. Thus we find a much larger gap between classical and quantum algorithms than the univariate case, where the speedup is by a factor of $2$. For the case of $\mathbb{F}_q$, we conjecture that $2k_{\mathbb{C}}$ queries also suffice to achieve probability approaching $1$ for large field order $q$, although we leave this as an open problem., Comment: 15 pages, 0 figures. Comments are welcome

Narrated 360{\deg} videos are typically provided in many touring scenarios to mimic real-world experience. However, previous work has shown that smart assistance (i.e., providing visual guidance) can significantly help users to follow the Normal Field of View (NFoV) corresponding to the narrative. In this project, we aim at automatically grounding the NFoVs of a 360{\deg} video given subtitles of the narrative (referred to as "NFoV-grounding"). We propose a novel Visual Grounding Model (VGM) to implicitly and efficiently predict the NFoVs given the video content and subtitles. Specifically, at each frame, we efficiently encode the panorama into feature map of candidate NFoVs using a Convolutional Neural Network (CNN) and the subtitles to the same hidden space using an RNN with Gated Recurrent Units (GRU). Then, we apply soft-attention on candidate NFoVs to trigger sentence decoder aiming to minimize the reconstruct loss between the generated and given sentence. Finally, we obtain the NFoV as the candidate NFoV with the maximum attention without any human supervision. To train VGM more robustly, we also generate a reverse sentence conditioning on one minus the soft-attention such that the attention focuses on candidate NFoVs less relevant to the given sentence. The negative log reconstruction loss of the reverse sentence (referred to as "irrelevant loss") is jointly minimized to encourage the reverse sentence to be different from the given sentence. To evaluate our method, we collect the first narrated 360{\deg} videos dataset and achieve state-of-the-art NFoV-grounding performance.

We consider the number of quantum queries required to determine the coefficients of a degree-d polynomial over GF(q). A lower bound shown independently by Kane and Kutin and by Meyer and Pommersheim shows that d/2+1/2 quantum queries are needed to solve this problem with bounded error, whereas an algorithm of Boneh and Zhandry shows that d quantum queries are sufficient. We show that the lower bound is achievable: d/2+1/2 quantum queries suffice to determine the polynomial with bounded error. Furthermore, we show that d/2+1 queries suffice to achieve probability approaching 1 for large q. These upper bounds improve results of Boneh and Zhandry on the insecurity of cryptographic protocols against quantum attacks. We also show that our algorithm's success probability as a function of the number of queries is precisely optimal. Furthermore, the algorithm can be implemented with gate complexity poly(log q) with negligible decrease in the success probability. We end with a conjecture about the quantum query complexity of multivariate polynomial interpolation., Comment: 17 pages, minor improvements, added conjecture about multivariate interpolation

Sample preparation is a significant challenge for detection and sensing technologies, since the presence of blood cells can interfere with the accuracy and reliability of virus detection at the nanoscale for point-of-care testing. To the best of our knowledge, there is not an existing on-chip virus isolation technology that does not use complex fluidic pumps. Here, we presented a lab-on-a-chip filter device to isolate plasma and viruses from unprocessed whole blood based on size exclusion without using a micropump. We demonstrated that viruses (eg, HIV) can be separated on a filter-based chip (2-µm pore size) from HIV-spiked whole blood at high recovery efficiencies of 89.9% ± 5.0%, 80.5% ± 4.3%, and 78.2% ± 3.8%, for viral loads of 1000, 10,000 and 100,000 copies/mL, respectively. Meanwhile, 81.7% ± 6.7% of red blood cells and 89.5% ± 2.4% of white blood cells were retained on 2 µm pore–sized filter microchips. We also tested these filter microchips with seven HIV-infected patient samples and observed recovery efficiencies ranging from 73.1% ± 8.3% to 82.5% ± 4.1%. These results are first steps towards developing disposable point-of-care diagnostics and monitoring devices for resource-constrained settings, as well as hospital and primary care settings., National Institutes of Health (U.S.) (NIH R01 A1081534), National Institutes of Health (U.S.) (NIH R01-A1093282), National Institutes of Health (U.S.) (R21 AI087107), National Institutes of Health (U.S.) (NIH U54EB015408), National Institutes of Health (U.S.) (R21 HL095960), United States. Army Medical Research and Materiel Command, United States. Army. Telemedicine and Advanced Technology Research Center

Background Herpes simplex virus 1 (HSV-1) and varicella zoster virus (VZV) cause extensive intra-ocular and neural infections in humans and are closely related to Felid herpes virus 1 (FeHV-1). We report the extent of intra-ocular replication and the extent and morphological aspects of neural replication during the acute and latent phases of FeHV-1 infection. Juvenile, SPF cats were inoculated with FeHV-1. Additional cats were used as negative controls. Cats were euthanized on days 6, 10, and 30 post-inoculation. Results FeHV-1 was isolated from the conjunctiva, cornea, uveal tract, retina, optic nerve, ciliary ganglion (CG), pterygopalatine ganglion (PTPG), trigeminal ganglion (TG), brainstem, visual cortex, cerebellum, and olfactory bulb of infected cats during the acute phase, but not the cranial cervical ganglion (CCG) and optic chiasm. Viral DNA was detected in all tissues during acute infection by a real-time quantitative PCR assay. On day 30, viral DNA was detected in all TG, all CCG, and 2 PTPG. Histologically mild inflammation and ganglion cell loss were noted within the TG during acute, but not latent infection. Using linear regression, a strong correlation existed between clinical score and day 30 viral DNA copy number within the TG. Conclusions The correlation between clinical score and day 30 viral DNA copy number suggests the severity of the acute clinical infection is related to the quantity of latent viral DNA. The histologic response was similar to that seen during HSV-1 or VZV infection. To the author’s knowledge this is the first report of FeHV-1 infection involving intraocular structures and autonomic ganglia.

ShuQi Wang,1 Dusan Sarenac,1 Michael H Chen,1 Shih-Han Huang,1 Francoise F Giguel,2 Daniel R Kuritzkes,3 Utkan Demirci1,41Bio-acoustic MEMS in Medicine Laboratory, Department of Medicine, Division of Biomedical Engineering, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA; 2Infectious Diseases Unit, Massachusetts General Hospital, Boston, MA, USA; 3Section of Retroviral Therapeutics, Brigham and Women's Hospital, Boston, MA, USA; 4Harvard-MIT Health Sciences and Technology, Cambridge, MA, USAAbstract: Sample preparation is a significant challenge for detection and sensing technologies, since the presence of blood cells can interfere with the accuracy and reliability of virus detection at the nanoscale for point-of-care testing. To the best of our knowledge, there is not an existing on-chip virus isolation technology that does not use complex fluidic pumps. Here, we presented a lab-on-a-chip filter device to isolate plasma and viruses from unprocessed whole blood based on size exclusion without using a micropump. We demonstrated that viruses (eg, HIV) can be separated on a filter-based chip (2-µm pore size) from HIV-spiked whole blood at high recovery efficiencies of 89.9% ± 5.0%, 80.5% ± 4.3%, and 78.2% ± 3.8%, for viral loads of 1000, 10,000 and 100,000 copies/mL, respectively. Meanwhile, 81.7% ± 6.7% of red blood cells and 89.5% ± 2.4% of white blood cells were retained on 2 µm pore–sized filter microchips. We also tested these filter microchips with seven HIV-infected patient samples and observed recovery efficiencies ranging from 73.1% ± 8.3% to 82.5% ± 4.1%. These results are first steps towards developing disposable point-of-care diagnostics and monitoring devices for resource-constrained settings, as well as hospital and primary care settings.Keywords: microchip, filtration, virus isolation, plasma separation, point-o