Search

In an agricultural field, plant phenotyping using object detection models is gaining attention. However, collecting the training data necessary to create generic and high-precision models is extremely challenging due to the difficulty of annotation and the diversity of domains. Furthermore, it is difficult to transfer training data across different crops, and although machine learning models effective for specific environments, conditions, or crops have been developed, they cannot be widely applied in actual fields. In this study, we propose a generative data augmentation method (D4) for vineyard shoot detection. D4 uses a pre-trained text-guided diffusion model based on a large number of original images culled from video data collected by unmanned ground vehicles or other means, and a small number of annotated datasets. The proposed method generates new annotated images with background information adapted to the target domain while retaining annotation information necessary for object detection. In addition, D4 overcomes the lack of training data in agriculture, including the difficulty of annotation and diversity of domains. We confirmed that this generative data augmentation method improved the mean average precision by up to 28.65% for the BBox detection task and the average precision by up to 13.73% for the keypoint detection task for vineyard shoot detection. Our generative data augmentation method D4 is expected to simultaneously solve the cost and domain diversity issues of training data generation in agriculture and improve the generalization performance of detection models.

Zero-point fluctuations in the background of a cosmic string provide an opportunity to study the effects of topology in quantum field theory. We use a scattering theory approach to compute quantum corrections to the energy density of a cosmic string, using the "ballpoint pen" and "flowerpot" models to allow for a nonzero string radius. For computational efficiency, we consider a massless field in $2+1$ dimensions. We show how to implement precise and unambiguous renormalization conditions in the presence of a deficit angle, and make use of Kontorovich-Lebedev techniques to rewrite the sum over angular momentum channels as an integral on the imaginary axis., Comment: 12 pages, 9 .eps figures

Fine-tuning large language models on private data for downstream applications poses significant privacy risks in potentially exposing sensitive information. Several popular community platforms now offer convenient distribution of a large variety of pre-trained models, allowing anyone to publish without rigorous verification. This scenario creates a privacy threat, as pre-trained models can be intentionally crafted to compromise the privacy of fine-tuning datasets. In this study, we introduce a novel poisoning technique that uses model-unlearning as an attack tool. This approach manipulates a pre-trained language model to increase the leakage of private data during the fine-tuning process. Our method enhances both membership inference and data extraction attacks while preserving model utility. Experimental results across different models, datasets, and fine-tuning setups demonstrate that our attacks significantly surpass baseline performance. This work serves as a cautionary note for users who download pre-trained models from unverified sources, highlighting the potential risks involved.

In distributed learning settings, models are iteratively updated with shared gradients computed from potentially sensitive user data. While previous work has studied various privacy risks of sharing gradients, our paper aims to provide a systematic approach to analyze private information leakage from gradients. We present a unified game-based framework that encompasses a broad range of attacks including attribute, property, distributional, and user disclosures. We investigate how different uncertainties of the adversary affect their inferential power via extensive experiments on five datasets across various data modalities. Our results demonstrate the inefficacy of solely relying on data aggregation to achieve privacy against inference attacks in distributed learning. We further evaluate five types of defenses, namely, gradient pruning, signed gradient descent, adversarial perturbations, variational information bottleneck, and differential privacy, under both static and adaptive adversary settings. We provide an information-theoretic view for analyzing the effectiveness of these defenses against inference from gradients. Finally, we introduce a method for auditing attribute inference privacy, improving the empirical estimation of worst-case privacy through crafting adversarial canary records.

Bayesian learning provides a unified skeleton to solve the electrophysiological source imaging task. From this perspective, existing source imaging algorithms utilize the Gaussian assumption for the observation noise to build the likelihood function for Bayesian inference. However, the electromagnetic measurements of brain activity are usually affected by miscellaneous artifacts, leading to a potentially non-Gaussian distribution for the observation noise. Hence the conventional Gaussian likelihood model is a suboptimal choice for the real-world source imaging task. In this study, we aim to solve this problem by proposing a new likelihood model which is robust with respect to non-Gaussian noises. Motivated by the robust maximum correntropy criterion, we propose a new improper distribution model concerning the noise assumption. This new noise distribution is leveraged to structure a robust likelihood function and integrated with hierarchical prior distributions to estimate source activities by variational inference. In particular, the score matching is adopted to determine the hyperparameters for the improper likelihood model. A comprehensive performance evaluation is performed to compare the proposed noise assumption to the conventional Gaussian model. Simulation results show that, the proposed method can realize more precise source reconstruction by designing known ground-truth. The real-world dataset also demonstrates the superiority of our new method with the visual perception task. This study provides a new backbone for Bayesian source imaging, which would facilitate its application using real-world noisy brain signal.

Leveraging empirical data is crucial in the development of accurate and reliable virtual models for the advancement of autonomous ship technologies and the optimization of port operations. This study presents an in-depth analysis of ship berthing and unberthing maneuvering characteristics by utilizing a comprehensive dataset encompassing the operation of a full-scale ship in diverse infrastructural and environmental conditions. Various statistical techniques and time-series analysis were employed to process and interpret the operational data. A systematic analysis was conducted on key performance variables, including approach speed, drift angles, turning motions, distance from obstacles, and actuator utilization. The results demonstrate significant discrepancies between the empirical data and the established maneuvering characteristics. These findings have the potential to significantly enhance the accuracy and reliability of conventional maneuvering models, such as the Mathematical Modeling Group (MMG) model, and improve the conditions used in captive model tests for the identification of maneuvering model parameters. Furthermore, these findings could inform the development of more robust autonomous berthing and unberthing algorithms and digital twins.

Recent breakthroughs in understanding the human brain have revealed its impressive ability to efficiently process and interpret human thoughts, opening up possibilities for intervening in brain signals. In this paper, we aim to develop a straightforward framework that uses other modalities, such as natural language, to translate the original dreamland. We present DreamConnect, employing a dual-stream diffusion framework to manipulate visually stimulated brain signals. By integrating an asynchronous diffusion strategy, our framework establishes an effective interface with human dreams, progressively refining their final imagery synthesis. Through extensive experiments, we demonstrate the method ability to accurately instruct human brain signals with high fidelity. Our project will be publicly available on https://github.com/Sys-Nexus/DreamConnect

This paper investigates quantum machine learning to optimize the beamforming in a multiuser multiple-input single-output downlink system. We aim to combine the power of quantum neural networks and the success of classical deep neural networks to enhance the learning performance. Specifically, we propose two hybrid quantum-classical neural networks to maximize the sum rate of a downlink system. The first one proposes a quantum neural network employing parameterized quantum circuits that follows a classical convolutional neural network. The classical neural network can be jointly trained with the quantum neural network or pre-trained leading to a fine-tuning transfer learning method. The second one designs a quantum convolutional neural network to better extract features followed by a classical deep neural network. Our results demonstrate the feasibility of the proposed hybrid neural networks, and reveal that the first method can achieve similar sum rate performance compared to a benchmark classical neural network with significantly less training parameters; while the second method can achieve higher sum rate especially in presence of many users still with less training parameters. The robustness of the proposed methods is verified using both software simulators and hardware emulators considering noisy intermediate-scale quantum devices.

We construct a four-dimensional spacetime using a three-dimensional contact manifold equipped with a degenerate metric. The degenerate metric is set to be compatible with the contact structure. The compatibility condition is defined in this paper. Our construction yields a Ricci tensor of a particularly simple form, which leads to a solution of the Einstein equation with a null dust and cosmic strings. The solution includes two arbitrary functions: the energy density of the null dust and the number density of the cosmic strings. When there exist the cosmic strings, the spacetime is of Petrov type D. Otherwise, the spacetime is conformally flat. For some simple matter densities, we examine the Einstein equation in detail., Comment: 25 pages, 3 figures

There is a growing demand for simulators for the research and development of maritime autonomous surface ships (MASS) and the approval of autonomous navigation algorithms. Simulators are used for purposes such as evaluation and training and are taken on various configurations accordingly. The ship maneuvering mathematical model used in such a simulator is an important element that characterizes the simulator. In this paper, we discuss the dynamic model of the hull and its position in the simulator that will be required for MASSs in the future. It also discusses guidelines for selecting an appropriate model, which has not been discussed extensively in previous studies. Finally, we discuss the functional requirements that simulators should have., Comment: 21 pages, 3 figures, Submitted to Journal of Marine Science and Technology

We consider a viscous compressible barotropic flow in the interval $[0,\pi]$ with homogeneous Dirichlet boundary conditions for the flow velocity and a constant rest state as initial data. Given two sufficiently close subintervals $I=[\alpha_1,\alpha_2]$ and $J=[\beta_1,\beta_2]$ of $(0,1)$, a nonempty open set $\omega \subset (1,\pi)$, and $T>0$, we construct an external force $f$ supported in $\omega$ acting on the momentum equation such that the corresponding flow map moves the fluid particles initially occupying $I$ exactly onto $J$ in time $T$., Comment: 16 pages

When an ideal insulator is cooled, four regimes of thermal conductivity are expected to emerge one after another. Two of these, the Ziman and the Poiseuille, are hydrodynamic regimes in which collision among phonons are mostly Normal. It has been difficult to observe them, save for a few insulators with high levels of isotopic and chemical purity. Our thermal transport measurements, covering four decades of temperatures between 0.1 K and 900 K, reveal that sapphire displays all four regimes, despite its isotopic impurity. In the Ziman regime, the thermal conductivity exponentially increases attaining an amplitude as large as 35,000 W/Km. We show that the peak thermal conductivity of ultra-pure, simple insulators, including diamond, silicon and solid helium, is set by a universal scaling depending on isotropic purity. The thermal conductivity of sapphire is an order of magnitude higher than what is expected by this scaling. We argue that this may be caused by the proximity of optical and acoustic phonon modes, as a consequence of the large number of atoms in the primitive cell.

We call two copulas tail equivalent if their first-order approximations in the tail coincide. As a special case, a copula is called tail symmetric if it is tail equivalent to the associated survival copula. We propose a novel measure and statistical test for tail equivalence. The proposed measure takes the value of zero if and only if the two copulas share a pair of tail order and tail order parameter in common. Moreover, taking the nature of these tail quantities into account, we design the proposed measure so that it takes a large value when tail orders are different, and a small value when tail order parameters are non-identical. We derive asymptotic properties of the proposed measure, and then propose a novel statistical test for tail equivalence. Performance of the proposed test is demonstrated in a series of simulation studies and empirical analyses of financial stock returns in the periods of the world financial crisis and the COVID-19 recession. Our empirical analysis reveals non-identical tail behaviors in different pairs of stocks, different parts of tails, and the two periods of recessions.

Wi-Fi channel measurements across different bands, e.g., sub-7-GHz and 60-GHz bands, are asynchronous due to the uncoordinated nature of distinct standards protocols, e.g., 802.11ac/ax/be and 802.11ad/ay. Multi-band Wi-Fi fusion has been considered before on a frame-to-frame basis for simple classification tasks, which does not require fine-time-scale alignment. In contrast, this paper considers asynchronous sequence-to-sequence fusion between sub-7-GHz channel state information (CSI) and 60-GHz beam signal-to-noise-ratio~(SNR)s for more challenging tasks such as continuous coordinate estimation. To handle the timing disparity between asynchronous multi-band Wi-Fi channel measurements, this paper proposes a multi-band neural dynamic fusion (NDF) framework. This framework uses separate encoders to embed the multi-band Wi-Fi measurement sequences to separate initial latent conditions. Using a continuous-time ordinary differential equation (ODE) modeling, these initial latent conditions are propagated to respective latent states of the multi-band channel measurements at the same time instances for a latent alignment and a post-ODE fusion, and at their original time instances for measurement reconstruction. We derive a customized loss function based on the variational evidence lower bound (ELBO) that balances between the multi-band measurement reconstruction and continuous coordinate estimation. We evaluate the NDF framework using an in-house multi-band Wi-Fi testbed and demonstrate substantial performance improvements over a comprehensive list of single-band and multi-band baseline methods., Comment: 13 pages, 13 figures, 4 tables

In-hand tool manipulation is an operation that not only manipulates a tool within the hand (i.e., in-hand manipulation) but also achieves a grasp suitable for a task after the manipulation. This study aims to achieve an in-hand tool manipulation skill through deep reinforcement learning. The difficulty of learning the skill arises because this manipulation requires (A) exploring long-term contact-state changes to achieve the desired grasp and (B) highly-varied motions depending on the contact-state transition. (A) leads to a sparsity of a reward on a successful grasp, and (B) requires an RL agent to explore widely within the state-action space to learn highly-varied actions, leading to sample inefficiency. To address these issues, this study proposes Action Primitives based on Contact-state Transition (APriCoT). APriCoT decomposes the manipulation into short-term action primitives by describing the operation as a contact-state transition based on three action representations (detach, crossover, attach). In each action primitive, fingers are required to perform short-term and similar actions. By training a policy for each primitive, we can mitigate the issues from (A) and (B). This study focuses on a fundamental operation as an example of in-hand tool manipulation: rotating an elongated object grasped with a precision grasp by half a turn to achieve the initial grasp. Experimental results demonstrated that ours succeeded in both the rotation and the achievement of the desired grasp, unlike existing studies. Additionally, it was found that the policy was robust to changes in object shape.

Randomized smoothing is a defensive technique to achieve enhanced robustness against adversarial examples which are small input perturbations that degrade the performance of neural network models. Conventional randomized smoothing adds random noise with a fixed noise level for every input sample to smooth out adversarial perturbations. This paper proposes a new variational framework that uses a per-sample noise level suitable for each input by introducing a noise level selector. Our experimental results demonstrate enhancement of empirical robustness against adversarial attacks. We also provide and analyze the certified robustness for our sample-wise smoothing method., Comment: 20 pages, under preparation

Biosignal-based hand gesture classification is an important component of effective human-machine interaction. For multimodal biosignal sensing, the modalities often face data loss due to missing channels in the data which can adversely affect the gesture classification performance. To make the classifiers robust to missing channels in the data, this paper proposes using Random Channel Ablation (RChA) during the training process. Ultrasound and force myography (FMG) data were acquired from the forearm for 12 hand gestures over 2 subjects. The resulting multimodal data had 16 total channels, 8 for each modality. The proposed method was applied to convolutional neural network architecture, and compared with baseline, imputation, and oracle methods. Using 5-fold cross-validation for the two subjects, on average, 12.2% and 24.5% improvement was observed for gesture classification with up to 4 and 8 missing channels respectively compared to the baseline. Notably, the proposed method is also robust to an increase in the number of missing channels compared to other methods. These results show the efficacy of using random channel ablation to improve classifier robustness for multimodal and multi-channel biosignal-based hand gesture classification., Comment: 5 pages, 4 figures

Large vision-language models (LVLMs), such as the Generative Pre-trained Transformer 4-omni (GPT-4o), are emerging multi-modal foundation models which have great potential as powerful artificial-intelligence (AI) assistance tools for a myriad of applications, including healthcare, industrial, and academic sectors. Although such foundation models perform well in a wide range of general tasks, their capability without fine-tuning is often limited in specialized tasks. However, full fine-tuning of large foundation models is challenging due to enormous computation/memory/dataset requirements. We show that GPT-4o can decode hand gestures from forearm ultrasound data even with no fine-tuning, and improves with few-shot, in-context learning., Comment: 8 pages, 9 figures

We investigate complex analytic properties of a neighborhood of a reduced rational curve with an ordinary cusp embedded in a non-singular complex surface whose self-intersection number is zero., Comment: 21 pages

As well-known, the $C^\infty$ stability of proper $C^\infty$ maps is characterized by the infinitesimal $C^\infty$ stability. In the present paper we study the counterpart in real analytic context. In particular, we show that the infinitesimal $C^\omega$ stability does not imply $C^\omega$ stability; for instance, a Whitney umbrella $\mathbb{R}^2 \to \mathbb{R}^3$ is not $C^\omega$ stable. A main tool for the proof is a relative version of Whitney's Analytic Approximation Theorem which is shown by using H. Cartan's Theorems A and B., Comment: 7 pages

The recent developments of Diffusion Models (DMs) enable generation of astonishingly high-quality synthetic samples. Recent work showed that the synthetic samples generated by the diffusion model, which is pre-trained on public data and fully fine-tuned with differential privacy on private data, can train a downstream classifier, while achieving a good privacy-utility tradeoff. However, fully fine-tuning such large diffusion models with DP-SGD can be very resource-demanding in terms of memory usage and computation. In this work, we investigate Parameter-Efficient Fine-Tuning (PEFT) of diffusion models using Low-Dimensional Adaptation (LoDA) with Differential Privacy. We evaluate the proposed method with the MNIST and CIFAR-10 datasets and demonstrate that such efficient fine-tuning can also generate useful synthetic samples for training downstream classifiers, with guaranteed privacy protection of fine-tuning data. Our source code will be made available on GitHub.

Traditional approaches to human-robot interaction design typically examine robot behaviors in controlled environments and narrow tasks. These methods are impractical for designing robots that interact with diverse user groups in complex human environments. Drawing from the field of theater, we present the construct of scenes -- individual environments consisting of specific people, objects, spatial arrangements, and social norms -- and tangible scenography, as a holistic design approach for human-robot interactions. We created a design tool, Tangible Scenography Kit (TaSK), with physical props to aid in design brainstorming. We conducted design sessions with eight professional designers to generate exploratory designs. Designers used tangible scenography and TaSK components to create multiple scenes with specific interaction goals, characterize each scene's social environment, and design scene-specific robot behaviors. From these sessions, we found that this method can encourage designers to think beyond a robot's narrow capabilities and consider how they can facilitate complex social interactions.

Text-conditioned image-to-video generation (TI2V) aims to synthesize a realistic video starting from a given image (e.g., a woman's photo) and a text description (e.g., "a woman is drinking water."). Existing TI2V frameworks often require costly training on video-text datasets and specific model designs for text and image conditioning. In this paper, we propose TI2V-Zero, a zero-shot, tuning-free method that empowers a pretrained text-to-video (T2V) diffusion model to be conditioned on a provided image, enabling TI2V generation without any optimization, fine-tuning, or introducing external modules. Our approach leverages a pretrained T2V diffusion foundation model as the generative prior. To guide video generation with the additional image input, we propose a "repeat-and-slide" strategy that modulates the reverse denoising process, allowing the frozen diffusion model to synthesize a video frame-by-frame starting from the provided image. To ensure temporal continuity, we employ a DDPM inversion strategy to initialize Gaussian noise for each newly synthesized frame and a resampling technique to help preserve visual details. We conduct comprehensive experiments on both domain-specific and open-domain datasets, where TI2V-Zero consistently outperforms a recent open-domain TI2V model. Furthermore, we show that TI2V-Zero can seamlessly extend to other tasks such as video infilling and prediction when provided with more images. Its autoregressive design also supports long video generation., Comment: CVPR 2024

The recent seminal work of Chernozhukov, Chetverikov and Kato has shown that bootstrap approximation for the maximum of a sum of independent random vectors is justified even when the dimension is much larger than the sample size. In this context, numerical experiments suggest that third-moment match bootstrap approximations would outperform normal approximation even without studentization, but the existing theoretical results cannot explain this phenomenon. In this paper, we first show that Edgeworth expansion, if justified, can give an explanation for this phenomenon. Second, we obtain valid Edgeworth expansions in the high-dimensional setting when the random vectors have Stein kernels. Finally, we prove the second-order accuracy of a double wild bootstrap method in this setting. As a byproduct, we find an interesting blessing of dimensionality phenomenon: The single third-moment match wild bootstrap is already second-order accurate in high-dimensions if the covariance matrix has identical diagonal entries and bounded eigenvalues., Comment: 61 pages, 1 figure, 2 tables. Some minor issues have been corrected

Sparse Bayesian learning has promoted many effective frameworks for brain activity decoding, especially for the reconstruction of muscle activity. However, existing sparse Bayesian learning mainly employs Gaussian distribution as error assumption in the reconstruction task, which is not necessarily the truth in the real-world application. On the other hand, brain recording is known to be highly noisy and contains many non-Gaussian noises, which could lead to significant performance degradation for sparse Bayesian learning method. The goal of this paper is to propose a new robust implementation for sparse Bayesian learning, so that robustness and sparseness can be realized simultaneously. Motivated by the great robustness of maximum correntropy criterion (MCC), we proposed an integration of MCC into the sparse Bayesian learning regime. To be specific, we derived the explicit error assumption inherent in the MCC and then leveraged it for the likelihood function. Meanwhile, we used the automatic relevance determination (ARD) technique for the sparse prior distribution. To fully evaluate the proposed method, a synthetic dataset and a real-world muscle activity reconstruction task with two different brain modalities were employed. Experimental results showed that our proposed sparse Bayesian correntropy learning framework improves significantly the robustness in a noisy regression task. The proposed method can realize higher correlation coefficient and lower root mean squared error in the real-world muscle activity reconstruction tasks. Sparse Bayesian correntropy learning provides a powerful tool for neural decoding which can promote the development of brain-computer interfaces.

Low-rank adaptation (LoRA) and its variants are widely employed in fine-tuning large models, including large language models for natural language processing and diffusion models for computer vision. This paper proposes a generalized framework called SuperLoRA that unifies and extends different LoRA variants, which can be realized under different hyper-parameter settings. Introducing grouping, folding, shuffling, projecting, and tensor factoring, SuperLoRA offers high flexibility compared with other LoRA variants and demonstrates superior performance for transfer learning tasks especially in the extremely few-parameter regimes., Comment: 33 pages, 29 figures

High-level synthesis (HLS) is a design flow that leverages modern language features and flexibility, such as complex data structures, inheritance, templates, etc., to prototype hardware designs rapidly. However, exploring various design space parameters can take much time and effort for hardware engineers to meet specific design specifications. This paper proposes a novel framework called AutoHLS, which integrates a deep neural network (DNN) with Bayesian optimization (BO) to accelerate HLS hardware design optimization. Our tool focuses on HLS pragma exploration and operation transformation. It utilizes integrated DNNs to predict synthesizability within a given FPGA resource budget. We also investigate the potential of emerging quantum neural networks (QNNs) instead of classical DNNs for the AutoHLS pipeline. Our experimental results demonstrate up to a 70-fold speedup in exploration time., Comment: 5 pages, 6 figures, MWSCAS 2023

While considerable progress has been made in achieving accurate lip synchronization for 3D speech-driven talking face generation, the task of incorporating expressive facial detail synthesis aligned with the speaker's speaking status remains challenging. Our goal is to directly leverage the inherent style information conveyed by human speech for generating an expressive talking face that aligns with the speaking status. In this paper, we propose AVI-Talking, an Audio-Visual Instruction system for expressive Talking face generation. This system harnesses the robust contextual reasoning and hallucination capability offered by Large Language Models (LLMs) to instruct the realistic synthesis of 3D talking faces. Instead of directly learning facial movements from human speech, our two-stage strategy involves the LLMs first comprehending audio information and generating instructions implying expressive facial details seamlessly corresponding to the speech. Subsequently, a diffusion-based generative network executes these instructions. This two-stage process, coupled with the incorporation of LLMs, enhances model interpretability and provides users with flexibility to comprehend instructions and specify desired operations or modifications. Extensive experiments showcase the effectiveness of our approach in producing vivid talking faces with expressive facial movements and consistent emotional status.

Large Language Models (LLMs) are widely used to evaluate natural language generation tasks as automated metrics. However, the likelihood, a measure of LLM's plausibility for a sentence, can vary due to superficial differences in sentences, such as word order and sentence structure. It is therefore possible that there might be a likelihood bias if LLMs are used for evaluation: they might overrate sentences with higher likelihoods while underrating those with lower likelihoods. In this paper, we investigate the presence and impact of likelihood bias in LLM-based evaluators. We also propose a method to mitigate the likelihood bias. Our method utilizes highly biased instances as few-shot examples for in-context learning. Our experiments in evaluating the data-to-text and grammatical error correction tasks reveal that several LLMs we test display a likelihood bias. Furthermore, our proposed method successfully mitigates this bias, also improving evaluation performance (in terms of correlation of models with human scores) significantly., Comment: 4 main pages

Let $X$ be a complex surface obtained as the quotient of the complex Euclidean space $\mathbb{C}^2$ by a discrete subgroup of rank $3$. We investigate the cohomology group $H_0^1(X, E)$ with compact support for a unitary flat line bundle $E$ over $X$. We show the vanishing of $H_0^1(X, E)$ for a certain class of such pairs $(X, E)$, which includes infinitely many examples such that $H^1(X, E)$ is non-Hausdorff and infinite dimensional., Comment: 17 pages

For small privacy parameter $\epsilon$, $\epsilon$-differential privacy (DP) provides a strong worst-case guarantee that no membership inference attack (MIA) can succeed at determining whether a person's data was used to train a machine learning model. The guarantee of DP is worst-case because: a) it holds even if the attacker already knows the records of all but one person in the data set; and b) it holds uniformly over all data sets. In practical applications, such a worst-case guarantee may be overkill: practical attackers may lack exact knowledge of (nearly all of) the private data, and our data set might be easier to defend, in some sense, than the worst-case data set. Such considerations have motivated the industrial deployment of DP models with large privacy parameter (e.g. $\epsilon \geq 7$), and it has been observed empirically that DP with large $\epsilon$ can successfully defend against state-of-the-art MIAs. Existing DP theory cannot explain these empirical findings: e.g., the theoretical privacy guarantees of $\epsilon \geq 7$ are essentially vacuous. In this paper, we aim to close this gap between theory and practice and understand why a large DP parameter can prevent practical MIAs. To tackle this problem, we propose a new privacy notion called practical membership privacy (PMP). PMP models a practical attacker's uncertainty about the contents of the private data. The PMP parameter has a natural interpretation in terms of the success rate of a practical MIA on a given data set. We quantitatively analyze the PMP parameter of two fundamental DP mechanisms: the exponential mechanism and Gaussian mechanism. Our analysis reveals that a large DP parameter often translates into a much smaller PMP parameter, which guarantees strong privacy against practical MIAs. Using our findings, we offer principled guidance for practitioners in choosing the DP parameter., Comment: Accepted at PPAI-24: AAAI Workshop on Privacy-Preserving Artificial Intelligence

This research incorporates realized volatility and overnight information into risk models, wherein the overnight return often contributes significantly to the total return volatility. Extending a semi-parametric regression model based on asymmetric Laplace distribution, we propose a family of RES-CAViaR-oc models by adding overnight return and realized measures as a nowcasting technique for simultaneously forecasting Value-at-Risk (VaR) and expected shortfall (ES). We utilize Bayesian methods to estimate unknown parameters and forecast VaR and ES jointly for the proposed model family. We also conduct extensive backtests based on joint elicitability of the pair of VaR and ES during the out-of sample period. Our empirical study on four international stock indices confirms that overnight return and realized volatility are vital in tail risk forecasting.

Capital allocation is a procedure for quantifying the contribution of each source of risk to aggregated risk. The gradient allocation rule, also known as the Euler principle, is a prevalent rule of capital allocation under which the allocated capital captures the diversification benefit of the marginal risk as a component of overall risk. This research concentrates on Expected Shortfall (ES) as a regulatory standard and focuses on the gradient allocations of ES, also called ES contributions (ESCs). We present the comprehensive treatment of backtesting the tuple of ESCs in the framework of the traditional and comparative backtests based on the concepts of joint identifiability and multi-objective elicitability. For robust forecast evaluation against the choice of scoring function, we also extend the Murphy diagram, a graphical tool to check whether one forecast dominates another under a class of scoring functions, to the case of ESCs. Finally, leveraging the recent concept of multi-objective elicitability, we propose a novel semiparametric model for forecasting dynamic ESCs based on a compositional regression model. In an empirical analysis of stock returns we evaluate and compare a variety of models for forecasting dynamic ESCs and demonstrate the outstanding performance of the proposed model.

In this paper, we explore how techniques from soft robotics can help create a new form of robot expression. We present Sprout, a soft expressive robot that conveys its internal states by changing its body shape. Sprout can extend, bend, twist, and expand using fiber-embedded actuators integrated into its construction. These deformations enable Sprout to express its internal states, for example, by expanding to express anger and bending its body sideways to express curiosity. Through two user studies, we investigated how users interpreted Sprout's expressions, their perceptions of Sprout, and their expectations from future iterations of Sprout's design. We argue that the use of soft actuators opens a novel design space for robot expressions to convey internal states, emotions, and intent., Comment: 10 pages, 5 figures

Many-query computations, in which a computational model for an engineering system must be evaluated many times, are crucial in design and control. For systems governed by partial differential equations (PDEs), typical high-fidelity numerical models are high-dimensional and too computationally expensive for the many-query setting. Thus, efficient surrogate models are required to enable low-cost computations in design and control. This work presents a physics-preserving reduced model learning approach that targets PDEs whose quadratic operators preserve energy, such as those arising in governing equations in many fluids problems. The approach is based on the Operator Inference method, which fits reduced model operators to state snapshot and time derivative data in a least-squares sense. However, Operator Inference does not generally learn a reduced quadratic operator with the energy-preserving property of the original PDE. Thus, we propose a new energy-preserving Operator Inference (EP-OpInf) approach, which imposes this structure on the learned reduced model via constrained optimization. Numerical results using the viscous Burgers' and Kuramoto-Sivashinksy equation (KSE) demonstrate that EP-OpInf learns efficient and accurate reduced models that retain this energy-preserving structure., Comment: 17 pages, AIAA SciTech Forum 2024