A novel single-lead handheld atrial fibrillation detection system

Objective: The single-lead handheld atrial fibrillation (AF) detection device is suitable for daily monitoring or early screening of AF in the hospital. However, the signal quality and the reliability of AF detection algorithm still need to be improved. This study proposed a novel AF detection system with a user-friendly interaction and a lightweight and accurate AF detection algorithm. Approach: The system consisted of a single-lead handheld electrocardiogram (ECG) device with a novel appearance like a gaming handle and a smartphone terminal embedded with AF detection. After feature optimization, the rule-based multi-feature AF detection algorithm had relatively good AF detection ability. Three types of experiments were designed to test the performance of the system. 1) Test the accuracy and time/memory cost of the AF detection algorithm. 2) Compare the proposed device with the standard device Shimmer. 3) Use the simulator to test the effectiveness of the system. Main results: The percentage of differences of successive RR intervals larger than 50 ms (PNN50), minimum value of RR intervals (minRR), and coefficient of sample entropy (COSEn) were features chosen for AF detection. 1) The sensitivity, specificity, and accuracy were 96.00%, 99.75%, 97.88% on the MIT-BIH AF database, and 98.50%, 94.50%, 96.50% on the clinical database we founded. The time/memory cost of the proposed algorithm was much smaller than that of Support Vector Machine (SVM). 2) The mean correlation coefficient of RR was 0.9950, indicating a high degree of consistency. 3) This system showed the effectiveness of AF detection. Significance: The proposed single-lead handheld AF detection system is demonstrated to be accurate, lightweight, consistent with the standard device, and efficient for AF detection.

Enhancing Urban Flow Maps via Neural ODEs

Flow super-resolution (FSR) enables inferring fine-grained urban flows with coarse-grained observations and plays an important role in traffic monitoring and prediction. The existing FSR solutions rely on deep CNN models (e.g., ResNet) for learning spatial correlation, incurring excessive memory cost and numerous parameter updates. We propose to tackle the urban flows inference using dynamic systems paradigm and present a new method FODE -- FSR with Ordinary Differential Equations (ODEs). FODE extends neural ODEs by introducing an affine coupling layer to overcome the problem of numerically unstable gradient computation, which allows more accurate and efficient spatial correlation estimation, without extra memory cost. In addition, FODE provides a flexible balance between flow inference accuracy and computational efficiency. A FODE-based augmented normalization mechanism is further introduced to constrain the flow distribution with the influence of external factors. Experimental evaluations on two real-world datasets demonstrate that FODE significantly outperforms several baseline approaches.

Hashing Based Answer Selection

Answer selection is an important subtask of question answering (QA), in which deep models usually achieve better performance than non-deep models. Most deep models adopt question-answer interaction mechanisms, such as attention, to get vector representations for answers. When these interaction based deep models are deployed for online prediction, the representations of all answers need to be recalculated for each question. This procedure is time-consuming for deep models with complex encoders like BERT which usually have better accuracy than simple encoders. One possible solution is to store the matrix representation (encoder output) of each answer in memory to avoid recalculation. But this will bring large memory cost. In this paper, we propose a novel method, called hashing based answer selection (HAS), to tackle this problem. HAS adopts a hashing strategy to learn a binary matrix representation for each answer, which can dramatically reduce the memory cost for storing the matrix representations of answers. Hence, HAS can adopt complex encoders like BERT in the model, but the online prediction of HAS is still fast with a low memory cost. Experimental results on three popular answer selection datasets show that HAS can outperform existing models to achieve state-of-the-art performance.

CAG: A Real-Time Low-Cost Enhanced-Robustness High-Transferability Content-Aware Adversarial Attack Generator

Deep neural networks (DNNs) are vulnerable to adversarial attack despite their tremendous success in many artificial intelligence fields. Adversarial attack is a method that causes the intended misclassfication by adding imperceptible perturbations to legitimate inputs. To date, researchers have developed numerous types of adversarial attack methods. However, from the perspective of practical deployment, these methods suffer from several drawbacks such as long attack generating time, high memory cost, insufficient robustness and low transferability. To address the drawbacks, we propose a Content-aware Adversarial Attack Generator (CAG) to achieve real-time, low-cost, enhanced-robustness and high-transferability adversarial attack. First, as a type of generative model-based attack, CAG shows significant speedup (at least 500 times) in generating adversarial examples compared to the state-of-the-art attacks such as PGD and C&W. Furthermore, CAG only needs a single generative model to perform targeted attack to any targeted class. Because CAG encodes the label information into a trainable embedding layer, it differs from prior generative model-based adversarial attacks that use n different copies of generative models for n different targeted classes. As a result, CAG significantly reduces the required memory cost for generating adversarial examples. Moreover, CAG can generate adversarial perturbations that focus on the critical areas of input by integrating the class activation maps information in the training process, and hence improve the robustness of CAG attack against the state-of-art adversarial defenses. In addition, CAG exhibits high transferability across different DNN classifier models in black-box attack scenario by introducing random dropout in the process of generating perturbations. Extensive experiments on different datasets and DNN models have verified the real-time, low-cost, enhanced-robustness, and high-transferability benefits of CAG.

Practical Key-Recovery Attacks On Round-Reduced Ketje Jr, Xoodoo-AE And Xoodyak

A new conditional cube attack was proposed by Li et al. at ToSC 2019 for cryptanalysis of Keccak keyed modes. In this paper, we find a new property of Li et al.’s method. The conditional cube attack is modified and applied to cryptanalysis of 5-round Ketje Jr, 6-round Xoodoo-AE and Xoodyak, where Ketje Jr is among the third round CAESAR competition candidates and Xoodyak is a Round 2 submission of the ongoing NIST lightweight cryptography project. For the updated conditional cube attack, all our results are shown to be of practical time complexity with negligible memory cost, and test codes are provided. Notably, our results on Xoodyak represent the first third-party cryptanalysis for Xoodyak.

Contextuality, memory cost and non-classicality for sequential measurements

The Kochen–Specker theorem, and the associated notion of quantum contextuality, can be considered as the starting point for the development of a notion of non-classical correlations for single systems. The subsequent debate around the possibility of an experimental test of Kochen–Specker-type contradiction stimulated the development of different theoretical frameworks to interpret experimental results. Starting from the approach based on sequential measurements, we will discuss a generalization of the notion of non-classical temporal correlations that goes beyond the contextuality approach and related ones based on Leggett and Garg's notion of macrorealism, and it is based on the notion of memory cost of generating correlations. Finally, we will review recent results on the memory cost for generating temporal correlations in classical and quantum systems. The present work is based on the talk given at the Purdue Winer Memorial Lectures 2018: probability and contextuality. This article is part of the theme issue ‘Contextuality and probability in quantum mechanics and beyond’.