Neural networks are seeing increased use in diverse Internet of Things (IoT) applications such as healthcare, smart homes and industrial monitoring. Their widespread use makes neural networks a lucrative target for theft. An attacker can obtain a model without having access to the training data or incurring the cost of training. Also, networks trained using private data (e.g., medical records) can reveal information about this data. Networks can be stolen by leveraging side channels such as power traces of the IoT device when it is running the network. Existing attacks require operations to occur in the same order each time; an attacker must collect and analyze several traces of the device to steal the network. Therefore, to prevent this type of attack, we randomly shuffle the order of operations each time. With shuffling, each operation can now happen at many different points in each execution, making the attack intractable. However, we show that shuffling in software can leak information which can be used to subvert this solution. Therefore, to perform secure shuffling and reduce latency, we present BlackJack, hardware added as a functional unit within the CPU. BlackJack secures neural networks on IoT devices by increasing the time needed for an attack to centuries, while adding just 2.46% area, 3.28% power and 0.56% latency overhead on an ARM M0+ SoC.
相關內容
Dynamical systems across the sciences, from electrical circuits to ecological networks, undergo qualitative and often catastrophic changes in behavior, called bifurcations, when their underlying parameters cross a threshold. Existing methods predict oncoming catastrophes in individual systems but are primarily time-series-based and struggle both to categorize qualitative dynamical regimes across diverse systems and to generalize to real data. To address this challenge, we propose a data-driven, physically-informed deep-learning framework for classifying dynamical regimes and characterizing bifurcation boundaries based on the extraction of topologically invariant features. We focus on the paradigmatic case of the supercritical Hopf bifurcation, which is used to model periodic dynamics across a wide range of applications. Our convolutional attention method is trained with data augmentations that encourage the learning of topological invariants which can be used to detect bifurcation boundaries in unseen systems and to design models of biological systems like oscillatory gene regulatory networks. We further demonstrate our method's use in analyzing real data by recovering distinct proliferation and differentiation dynamics along pancreatic endocrinogenesis trajectory in gene expression space based on single-cell data. Our method provides valuable insights into the qualitative, long-term behavior of a wide range of dynamical systems, and can detect bifurcations or catastrophic transitions in large-scale physical and biological systems.
Can generative AI help us speed up the authoring of tools to help self-represented litigants? In this paper, we describe 3 approaches to automating the completion of court forms: a generative AI approach that uses GPT-3 to iteratively prompt the user to answer questions, a constrained template-driven approach that uses GPT-4-turbo to generate a draft of questions that are subject to human review, and a hybrid method. We use the open source Docassemble platform in all 3 experiments, together with a tool created at Suffolk University Law School called the Assembly Line Weaver. We conclude that the hybrid model of constrained automated drafting with human review is best suited to the task of authoring guided interviews.
The recent success of neural networks in natural language processing has drawn renewed attention to learning sequence-to-sequence (seq2seq) tasks. While there exists a rich literature that studies classification and regression tasks using solvable models of neural networks, seq2seq tasks have not yet been studied from this perspective. Here, we propose a simple model for a seq2seq task that has the advantage of providing explicit control over the degree of memory, or non-Markovianity, in the sequences -- the stochastic switching-Ornstein-Uhlenbeck (SSOU) model. We introduce a measure of non-Markovianity to quantify the amount of memory in the sequences. For a minimal auto-regressive (AR) learning model trained on this task, we identify two learning regimes corresponding to distinct phases in the stationary state of the SSOU process. These phases emerge from the interplay between two different time scales that govern the sequence statistics. Moreover, we observe that while increasing the integration window of the AR model always improves performance, albeit with diminishing returns, increasing the non-Markovianity of the input sequences can improve or degrade its performance. Finally, we perform experiments with recurrent and convolutional neural networks that show that our observations carry over to more complicated neural network architectures.
We explore the power of the unbounded Fan-Out gate and the Global Tunable gates generated by Ising-type Hamiltonians in constructing constant-depth quantum circuits, with particular attention to quantum memory devices. We propose two types of constant-depth constructions for implementing Uniformly Controlled Gates. These gates include the Fan-In gates defined by $|x\rangle|b\rangle\mapsto |x\rangle|b\oplus f(x)\rangle$ for $x\in\{0,1\}^n$ and $b\in\{0,1\}$, where $f$ is a Boolean function. The first of our constructions is based on computing the one-hot encoding of the control register $|x\rangle$, while the second is based on Boolean analysis and exploits different representations of $f$ such as its Fourier expansion. Via these constructions, we obtain constant-depth circuits for the quantum counterparts of read-only and read-write memory devices -- Quantum Random Access Memory (QRAM) and Quantum Random Access Gate (QRAG) -- of memory size $n$. The implementation based on one-hot encoding requires either $O(n\log{n}\log\log{n})$ ancillae and $O(n\log{n})$ Fan-Out gates or $O(n\log{n})$ ancillae and $6$ Global Tunable gates. On the other hand, the implementation based on Boolean analysis requires only $2$ Global Tunable gates at the expense of $O(n^2)$ ancillae.
Intrusion Detection Systems (IDS) are widely employed to detect and mitigate external network security events. VANETs (Vehicle ad-hoc Networks) are evolving, especially with the development of Connected Autonomous Vehicles (CAVs). So, it is crucial to assess how traditional IDS approaches can be utilised for emerging technologies. To address this concern, our work presents a stacked ensemble learning approach for IDS, which combines multiple machine learning algorithms to detect threats more effectively than single algorithm methods. Using the CICIDS2017 and the VeReMi benchmark data sets, we compare the performance of our approach with existing machine learning methods and find that it is more accurate at identifying threats. Our method also incorporates hyperparameter optimization and feature selection to improve its performance further. Overall, our results suggest that stacked ensemble learning is a promising technique for enhancing the effectiveness of IDS.
IoT devices trigger real-time applications by receiving data from their vicinity. Modeling these applications in the form of workflows enables automating their procedure, especially for the business and industry. Depending on the features of the applications, they can be modeled in different forms, including single workflow, multiple workflows, and workflow ensembles. Since the whole data must be sent to the cloud servers for processing and storage, cloud computing has many challenges for executing real-time applications, such as bandwidth limitation, delay, and privacy. Edge paradigms are introduced to address the challenges of cloud computing in executing IoT applications. Executing IoT applications using device-to-device communications in edge paradigms requiring direct communication between devices in a network with a graph topology. While there is no simulator supporting simulating workflow-based applications and device-to-device communication, this paper introduces a toolkit for simulating resource management of scientific workflows in distributed environments with graph topology called WIDESim.The graph topology of WIDESim enables D2D communications in edge paradigms. WIDESim can work with all three different structures of scientific workflows: single, multiple workflows, and workflow ensembles. It has no constraint on the topology of the distributed environment. Also, unlike most existing network simulators, this simulator enables dynamic resource management and scheduling. We have validated the performance of WIDESim in comparison to standard simulators and workflow management tools. Also, we have evaluated its performance in different scenarios of distributed computing systems using different types of workflow-based applications. The results indicate that WIDESim's performance is close to existing standard simulators besides its improvements.
Variational autoencoder (VAE) and generative adversarial networks (GAN) have found widespread applications in clustering and have achieved significant success. However, the potential of these approaches may be limited due to VAE's mediocre generation capability or GAN's well-known instability during adversarial training. In contrast, denoising diffusion probabilistic models (DDPMs) represent a new and promising class of generative models that may unlock fresh dimensions in clustering. In this study, we introduce an innovative expectation-maximization (EM) framework for clustering using DDPMs. In the E-step, we aim to derive a mixture of Gaussian priors for the subsequent M-step. In the M-step, our focus lies in learning clustering-friendly latent representations for the data by employing the conditional DDPM and matching the distribution of latent representations to the mixture of Gaussian priors. We present a rigorous theoretical analysis of the optimization process in the M-step, proving that the optimizations are equivalent to maximizing the lower bound of the Q function within the vanilla EM framework under certain constraints. Comprehensive experiments validate the advantages of the proposed framework, showcasing superior performance in clustering, unsupervised conditional generation and latent representation learning.
We present a modification to RingCT protocol with stealth addresses that makes it compatible with Delegated Proof of Stake based consensus mechanisms called Delegated RingCT. Our scheme has two building blocks: a customised version of an Integrated Signature and Encryption scheme composed of a public key encryption scheme and two signature schemes (a digital signature and a linkable ring signature); and non-interactive zero knowledge proofs. We give a description of the scheme, security proofs and a prototype implementation whose benchmarking is discussed. Although Delegated RingCT doesn't have the same degree of anonymity as other RingCT constructions, we argue that the benefits that the compatibility with DPoS consensus mechanisms brings constitutes a reasonable trade-off for being able to develop an anonymous decentralised cryptocurrency that is faster and more scalable than existing ones.
We present ResMLP, an architecture built entirely upon multi-layer perceptrons for image classification. It is a simple residual network that alternates (i) a linear layer in which image patches interact, independently and identically across channels, and (ii) a two-layer feed-forward network in which channels interact independently per patch. When trained with a modern training strategy using heavy data-augmentation and optionally distillation, it attains surprisingly good accuracy/complexity trade-offs on ImageNet. We will share our code based on the Timm library and pre-trained models.
Recent advances in 3D fully convolutional networks (FCN) have made it feasible to produce dense voxel-wise predictions of volumetric images. In this work, we show that a multi-class 3D FCN trained on manually labeled CT scans of several anatomical structures (ranging from the large organs to thin vessels) can achieve competitive segmentation results, while avoiding the need for handcrafting features or training class-specific models. To this end, we propose a two-stage, coarse-to-fine approach that will first use a 3D FCN to roughly define a candidate region, which will then be used as input to a second 3D FCN. This reduces the number of voxels the second FCN has to classify to ~10% and allows it to focus on more detailed segmentation of the organs and vessels. We utilize training and validation sets consisting of 331 clinical CT images and test our models on a completely unseen data collection acquired at a different hospital that includes 150 CT scans, targeting three anatomical organs (liver, spleen, and pancreas). In challenging organs such as the pancreas, our cascaded approach improves the mean Dice score from 68.5 to 82.2%, achieving the highest reported average score on this dataset. We compare with a 2D FCN method on a separate dataset of 240 CT scans with 18 classes and achieve a significantly higher performance in small organs and vessels. Furthermore, we explore fine-tuning our models to different datasets. Our experiments illustrate the promise and robustness of current 3D FCN based semantic segmentation of medical images, achieving state-of-the-art results. Our code and trained models are available for download: //github.com/holgerroth/3Dunet_abdomen_cascade.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191