深度学习的最新进展，尤其是编码器架构的发明，已大大改善了抽象性摘要系统的性能。尽管大多数研究都集中在书面文件上，但我们观察到过去几年对对话和多方对话的总结越来越兴趣。一个可以可靠地将人类对话的音频或笔录转换为删节版本的系统，该版本在讨论中最重要的一点上可以在各种现实世界中，从商务会议到医疗咨询再到客户都有价值服务电话。本文着重于多党会议的抽象性摘要，对与此任务相关的挑战，数据集和系统进行了调查，并讨论了未来研究的有希望的方向。

Cataloging the complex behaviors of dynamical systems can be challenging, even when they are well-described by a simple mechanistic model. If such a system is of limited analytical tractability, brute force simulation is often the only resort. We present an alternative, optimization-driven approach using tools from machine learning. We apply this approach to a novel, fully-optimizable, reaction-diffusion model which incorporates complex chemical reaction networks (termed "Dense Reaction-Diffusion Network" or "Dense RDN"). This allows us to systematically identify new states and behaviors, including pattern formation, dissipation-maximizing nonequilibrium states, and replication-like dynamical structures.

Many dialogue systems (DSs) lack characteristics humans have, such as emotion perception, factuality, and informativeness. Enhancing DSs with knowledge alleviates this problem, but, as many ways of doing so exist, keeping track of all proposed methods is difficult. Here, we present the first survey of knowledge-enhanced DSs. We define three categories of systems - internal, external, and hybrid - based on the knowledge they use. We survey the motivation for enhancing DSs with knowledge, used datasets, and methods for knowledge search, knowledge encoding, and knowledge incorporation. Finally, we propose how to improve existing systems based on theories from linguistics and cognitive science.

Scene understanding is a major challenge of today's computer vision. Center to this task is image segmentation, since scenes are often provided as a set of pictures. Nowadays, many such datasets also provide 3D geometry information given as a 3D point cloud acquired by a laser scanner or a depth camera. To exploit this geometric information, many current approaches rely on both a 2D loss and 3D loss, requiring not only 2D per pixel labels but also 3D per point labels. However obtaining a 3D groundtruth is challenging, time-consuming and error-prone. In this paper, we show that image segmentation can benefit from 3D geometric information without requiring any 3D groundtruth, by training the geometric feature extraction with a 2D segmentation loss in an end-to-end fashion. Our method starts by extracting a map of 3D features directly from the point cloud by using a lightweight and simple 3D encoder neural network. The 3D feature map is then used as an additional input to a classical image segmentation network. During training, the 3D features extraction is optimized for the segmentation task by back-propagation through the entire pipeline. Our method exhibits state-of-the-art performance with much lighter input dataset requirements, since no 3D groundtruth is required.

In the upcoming years, artificial intelligence (AI) is going to transform the practice of medicine in most of its specialties. Deep learning can help achieve better and earlier problem detection, while reducing errors on diagnosis. By feeding a deep neural network (DNN) with the data from a low-cost and low-accuracy sensor array, we demonstrate that it becomes possible to significantly improve the measurements' precision and accuracy. The data collection is done with an array composed of 32 temperature sensors, including 16 analog and 16 digital sensors. All sensors have accuracies between 0.5-2.0$^\circ$C. 800 vectors are extracted, covering a range from to 30 to 45$^\circ$C. In order to improve the temperature readings, we use machine learning to perform a linear regression analysis through a DNN. In an attempt to minimize the model's complexity in order to eventually run inferences locally, the network with the best results involves only three layers using the hyperbolic tangent activation function and the Adam Stochastic Gradient Descent (SGD) optimizer. The model is trained with a randomly-selected dataset using 640 vectors (80% of the data) and tested with 160 vectors (20%). Using the mean squared error as a loss function between the data and the model's prediction, we achieve a loss of only 1.47x10$^{-4}$ on the training set and 1.22x10$^{-4}$ on the test set. As such, we believe this appealing approach offers a new pathway towards significantly better datasets using readily-available ultra low-cost sensors.

Machine learning models are frequently employed to perform either purely physics-free or hybrid downscaling of climate data. However, the majority of these implementations operate over relatively small downscaling factors of about 4--6x. This study examines the ability of convolutional neural networks (CNN) to downscale surface wind speed data from three different coarse resolutions (25km, 48km, and 100km side-length grid cells) to 3km and additionally focuses on the ability to recover subgrid-scale variability. Within each downscaling factor, namely 8x, 16x, and 32x, we consider models that produce fine-scale wind speed predictions as functions of different input features: coarse wind fields only; coarse wind and fine-scale topography; and coarse wind, topography, and temporal information in the form of a timestamp. Furthermore, we train one model at 25km to 3km resolution whose fine-scale outputs are probability density function parameters through which sample wind speeds can be generated. All CNN predictions performed on one out-of-sample data outperform classical interpolation. Models with coarse wind and fine topography are shown to exhibit the best performance compared to other models operating across the same downscaling factor. Our timestamp encoding results in lower out-of-sample generalizability compared to other input configurations. Overall, the downscaling factor plays the largest role in model performance.

Quantum-enhanced data science, also known as quantum machine learning (QML), is of growing interest as an application of near-term quantum computers. Variational QML algorithms have the potential to solve practical problems on real hardware, particularly when involving quantum data. However, training these algorithms can be challenging and calls for tailored optimization procedures. Specifically, QML applications can require a large shot-count overhead due to the large datasets involved. In this work, we advocate for simultaneous random sampling over both the dataset as well as the measurement operators that define the loss function. We consider a highly general loss function that encompasses many QML applications, and we show how to construct an unbiased estimator of its gradient. This allows us to propose a shot-frugal gradient descent optimizer called Refoqus (REsource Frugal Optimizer for QUantum Stochastic gradient descent). Our numerics indicate that Refoqus can save several orders of magnitude in shot cost, even relative to optimizers that sample over measurement operators alone.

Banks hold a societal responsibility and regulatory requirements to mitigate the risk of financial crimes. Risk mitigation primarily happens through monitoring customer activity through Transaction Monitoring (TM). Recently, Machine Learning (ML) has been proposed to identify suspicious customer behavior, which raises complex socio-technical implications around trust and explainability of ML models and their outputs. However, little research is available due to its sensitivity. We aim to fill this gap by presenting empirical research exploring how ML supported automation and augmentation affects the TM process and stakeholders' requirements for building eXplainable Artificial Intelligence (xAI). Our study finds that xAI requirements depend on the liable party in the TM process which changes depending on augmentation or automation of TM. Context-relatable explanations can provide much-needed support for auditing and may diminish bias in the investigator's judgement. These results suggest a use case-specific approach for xAI to adequately foster the adoption of ML in TM.

自动语音识别（ASR）系统的转录质量在转录来自看不见的域的音频时会大大降低。我们提出了一种无监督的误差校正方法，用于无监督的ASR域适应性，旨在恢复域不匹配引起的转录误差。与依靠转录音频进行训练的现有校正方法不同，我们的方法仅需要针对目标域的未标记数据，在该数据中，将伪标记技术应用于生成校正培训样品。为了减少对伪数据的过度拟合，我们还提出了一个编码器校正模型，该模型可以考虑其他信息，例如对话上下文和声学特征。实验结果表明，我们的方法在未适应的ASR系统中获得了显着的单词错误率（WER）。校正模型也可以在其他适应方法的基础上应用，以相对额外的改善。

对自主导航和室内应用程序勘探机器人的最新兴趣刺激了对室内同时定位和映射（SLAM）机器人系统的研究。尽管大多数这些大满贯系统使用视觉和激光雷达传感器与探针传感器同时使用，但这些探针传感器会随着时间的流逝而漂移。为了打击这种漂移，视觉大满贯系统部署计算和内存密集型搜索算法来检测“环闭合”，这使得轨迹估计在全球范围内保持一致。为了绕过这些资源（计算和内存）密集算法，我们提出了VIWID，该算法将WiFi和视觉传感器集成在双层系统中。这种双层方法将局部和全局轨迹估计的任务分开，从而使VIWID资源有效，同时实现PAR或更好的性能到最先进的视觉大满贯。我们在四个数据集上展示了VIWID的性能，涵盖了超过1500 m的遍历路径，并分别显示出4.3倍和4倍的计算和记忆消耗量与最先进的视觉和LIDAR SLAM SLAM系统相比，具有PAR SLAM性能。