Searcharxiv⌕ Search

arXiv subjects

Kevin Smith

Publications and source records attributed to Kevin Smith.

52 records · Page 3Linked to original sources
arXiv2022

PSL is Dead. Long Live PSL

Property Specification Language (PSL) is a form of temporal logic that has been mainly used in discrete domains (e.g. formal hardware verification). In this paper, we show that by merging machine learning techniques with PSL monitors, we can extend PSL to work on continuous domains. We apply this technique in machine learning-based anomaly detection to analyze scenarios of real-time streaming events from continuous variables in order to detect abnormal behaviors of a system. By using machine learning with formal models, we leverage the strengths of both machine learning methods and formal semantics of time. On one hand, machine learning techniques can produce distributions on continuous variables, where abnormalities can be captured as deviations from the distributions. On the other hand, formal methods can characterize discrete temporal behaviors and relations that cannot be easily learned by machine learning techniques. Interestingly, the anomalies detected by machine learning and the underlying time representation used are discrete events. We implemented a temporal monitoring package (TEF) that operates in conjunction with normal data science packages for anomaly detection machine learning systems, and we show that TEF can be used to perform accurate interpretation of temporal correlation between events.

cs.LG↗
arXiv2021

Privacy Leakage over Dependent Attributes in One-Sided Differential Privacy

Providing a provable privacy guarantees while maintaining the utility of data is a challenging task in many real-world applications. Recently, a new framework called One-Sided Differential Privacy (OSDP) was introduced that extends existing differential privacy approaches. OSDP increases the utility of the data by taking advantage of the fact that not all records are sensitive. However, the previous work assumed that all records are statistically independent from each other. Motivated by occupancy data in building management systems, this paper extends the existing one-sided differential privacy framework. In this paper, we quantify the overall privacy leakage when the adversary is given dependency information between the records. In addition, we show how an optimization problem can be constructed that efficiently trades off between the utility and privacy.

cs.CR↗
arXiv2021

On the Hessian-cscK equations

In this paper, we propose a coupled system of complex Hessian equations which generalizes the equation for constant scalar curvature Kähler (cscK) metrics. We show this system can be realized variationally as the Euler-Lagrange equation of a Hessian version of the Mabuchi K-energy in an infinite dimensional space of $k$-Hessian potentials, which can be seen as an infinite dimensional Riemannian manifold with negative sectional curvature. Finally, we prove an a priori $C^0$-estimate for this system which depends on the Entropy, which generalizes a fundamental result of Chen and Cheng for cscK metrics.

math.DG↗
arXiv2021

Conditional De-Identification of 3D Magnetic Resonance Images

Privacy protection of medical image data is challenging. Even if metadata is removed, brain scans are vulnerable to attacks that match renderings of the face to facial image databases. Solutions have been developed to de-identify diagnostic scans by obfuscating or removing parts of the face. However, these solutions either fail to reliably hide the patient's identity or are so aggressive that they impair further analyses. We propose a new class of de-identification techniques that, instead of removing facial features, remodels them. Our solution relies on a conditional multi-scale GAN architecture. It takes a patient's MRI scan as input and generates a 3D volume conditioned on the patient's brain, which is preserved exactly, but where the face has been de-identified through remodeling. We demonstrate that our approach preserves privacy far better than existing techniques, without compromising downstream medical analyses. Analyses were run on the OASIS-3 and ADNI corpora.

eess.IV↗
arXiv2021

Is it Time to Replace CNNs with Transformers for Medical Images?

Convolutional Neural Networks (CNNs) have reigned for a decade as the de facto approach to automated medical image diagnosis. Recently, vision transformers (ViTs) have appeared as a competitive alternative to CNNs, yielding similar levels of performance while possessing several interesting properties that could prove beneficial for medical imaging tasks. In this work, we explore whether it is time to move to transformer-based models or if we should keep working with CNNs - can we trivially switch to transformers? If so, what are the advantages and drawbacks of switching to ViTs for medical image diagnosis? We consider these questions in a series of experiments on three mainstream medical image datasets. Our findings show that, while CNNs perform better when trained from scratch, off-the-shelf vision transformers using default hyperparameters are on par with CNNs when pretrained on ImageNet, and outperform their CNN counterparts when pretrained using self-supervision.

cs.CV↗
arXiv2021

Unsupervised Discovery of 3D Physical Objects from Video

We study the problem of unsupervised physical object discovery. While existing frameworks aim to decompose scenes into 2D segments based off each object's appearance, we explore how physics, especially object interactions, facilitates disentangling of 3D geometry and position of objects from video, in an unsupervised manner. Drawing inspiration from developmental psychology, our Physical Object Discovery Network (POD-Net) uses both multi-scale pixel cues and physical motion cues to accurately segment observable and partially occluded objects of varying sizes, and infer properties of those objects. Our model reliably segments objects on both synthetic and real scenes. The discovered object properties can also be used to reason about physical events.

cs.CV↗
arXiv2020

Decoupling Inherent Risk and Early Cancer Signs in Image-based Breast Cancer Risk Models

The ability to accurately estimate risk of developing breast cancer would be invaluable for clinical decision-making. One promising new approach is to integrate image-based risk models based on deep neural networks. However, one must take care when using such models, as selection of training data influences the patterns the network will learn to identify. With this in mind, we trained networks using three different criteria to select the positive training data (i.e. images from patients that will develop cancer): an inherent risk model trained on images with no visible signs of cancer, a cancer signs model trained on images containing cancer or early signs of cancer, and a conflated model trained on all images from patients with a cancer diagnosis. We find that these three models learn distinctive features that focus on different patterns, which translates to contrasts in performance. Short-term risk is best estimated by the cancer signs model, whilst long-term risk is best estimated by the inherent risk model. Carelessly training with all images conflates inherent risk with early cancer signs, and yields sub-optimal estimates in both regimes. As a consequence, conflated models may lead physicians to recommend preventative action when early cancer signs are already visible.

eess.IV↗
arXiv2020

Adding Seemingly Uninformative Labels Helps in Low Data Regimes

Evidence suggests that networks trained on large datasets generalize well not solely because of the numerous training examples, but also class diversity which encourages learning of enriched features. This raises the question of whether this remains true when data is scarce - is there an advantage to learning with additional labels in low-data regimes? In this work, we consider a task that requires difficult-to-obtain expert annotations: tumor segmentation in mammography images. We show that, in low-data settings, performance can be improved by complementing the expert annotations with seemingly uninformative labels from non-expert annotators, turning the task into a multi-class problem. We reveal that these gains increase when less expert data is available, and uncover several interesting properties through further studies. We demonstrate our findings on CSAW-S, a new dataset that we introduce here, and confirm them on two public datasets.

cs.CV↗
arXiv2020

Explanation-based Weakly-supervised Learning of Visual Relations with Graph Networks

Visual relationship detection is fundamental for holistic image understanding. However, the localization and classification of (subject, predicate, object) triplets remain challenging tasks, due to the combinatorial explosion of possible relationships, their long-tailed distribution in natural images, and an expensive annotation process. This paper introduces a novel weakly-supervised method for visual relationship detection that relies on minimal image-level predicate labels. A graph neural network is trained to classify predicates in images from a graph representation of detected objects, implicitly encoding an inductive bias for pairwise relations. We then frame relationship detection as the explanation of such a predicate classifier, i.e. we obtain a complete relation by recovering the subject and object of a predicted predicate. We present results comparable to recent fully- and weakly-supervised methods on three diverse and challenging datasets: HICO-DET for human-object interaction, Visual Relationship Detection for generic object-to-object relations, and UnRel for unusual triplets; demonstrating robustness to non-comprehensive annotations and good few-shot generalization.

cs.CV↗
arXiv2020

Parabolic complex Monge-Ampère equations on compact Hermitian manifolds

We prove the long-time existence and convergence of solutions to a general class of parabolic equations, not necessarily concave in the Hessian of the unknown function, on a compact Hermitian manifold. The limiting function is identified as the solution of an elliptic complex Monge-Ampère equation.

math.AP↗
arXiv2019

An empirical study of the relation between network architecture and complexity

In this preregistration submission, we propose an empirical study of how networks handle changes in complexity of the data. We investigate the effect of network capacity on generalization performance in the face of increasing data complexity. For this, we measure the generalization error for an image classification task where the number of classes steadily increases. We compare a number of modern architectures at different scales in this setting. The methodology, setup, and hypotheses described in this proposal were evaluated by peer review before experiments were conducted.

cs.LG↗
arXiv2018

Bayesian Uncertainty Estimation for Batch Normalized Deep Networks

We show that training a deep network using batch normalization is equivalent to approximate inference in Bayesian models. We further demonstrate that this finding allows us to make meaningful estimates of the model uncertainty using conventional architectures, without modifications to the network or the training procedure. Our approach is thoroughly validated by measuring the quality of uncertainty in a series of empirical experiments on different tasks. It outperforms baselines with strong statistical significance, and displays competitive performance with recent Bayesian approaches.

stat.ML↗
arXiv2017

Single Ion Imaging and Fluorescence Collection with a Parabolic Mirror Trap

Single trapped ion qubit is an excellent candidate for quantum computation and information, with additional ability to coherently couple to single photons. Efficient fluorescence collection is the most challenging part in remote entangled ion qubit state generation. To address this issue, we developed an ion trap combining a reflective parabolic surface with trap electrodes. This parabolic trap design covers a solid angle of 2pi steradians, and allows precise ion placement at the focal point of the parabola. We measured approximately 39% fluorescence collection from a single ion with this mirror, and analyzed the mirror optical performance. We observed single ion image spot size of about 3.4 times diffraction limit, improved to 2.8 times diffraction limit with the help of an external deformable mirror. The micromotion of ion is determined to be the limiting factor, and the result is consistent with theoretical calculation.

physics.ins-det↗
arXiv2016

New Generation Value Networks for Content Delivery

In this paper we paint a broad picture of the Internet content delivery market, by taking into consideration both economical and technical challenges that might drive the interactions among the stakeholders in the future. We focus on a few disrupting factors, namely ubiquitous encryption, traffic boost, network scalability, latency needs and network control; and try to figure out whether the current model (CDN) is robust against their variation, which optimization can be envisioned and how the most accredited option (ICN) can be of help.

cs.NI↗
arXiv2015

High-gradient High-charge CW Superconducting RF gun with CsK2Sb photocathode

High-gradient CW photo-injectors operating at high accelerating gradients promise to revolutionize many sciences and applications. They can establish the basis for super-bright monochromatic X-ray free-electron lasers, super-bright hadron beams, nuclear- waste transmutation or a new generation of microchip production. In this letter we report on our operation of a superconducting RF electron gun with a record-high accelerating gradient at the CsK2Sb photocathode (i.e. ~ 20 MV/m) generating a record-high bunch charge (i.e., 3 nC). We briefly describe the system and then detail our experimental results. This achievement opens new era in generating high-power electron beams with a very high brightness.

physics.acc-ph↗