Recently, dense passage retrieval has become a mainstream approach to finding relevant information in various natural language processing tasks. A number of studies have been devoted to improving the widely adopted dual-encoder architecture. However, most of the previous studies only consider query-centric similarity relation when learning the dual-encoder retriever. In order to capture more comprehensive similarity relations, we propose a novel approach that leverages both query-centric and PAssage-centric sImilarity Relations (called PAIR) for dense passage retrieval. To implement our approach, we make three major technical contributions by introducing formal formulations of the two kinds of similarity relations, generating high-quality pseudo labeled data via knowledge distillation, and designing an effective two-stage training procedure that incorporates passage-centric similarity relation constraint. Extensive experiments show that our approach significantly outperforms previous state-of-the-art models on both MSMARCO and Natural Questions datasets.

Over-parameterized deep networks trained using gradient-based optimizers are a popular choice for solving classification and ranking problems. Without appropriately tuned $\ell_2$ regularization or weight decay, such networks have the tendency to make output scores (logits) and network weights large, causing training loss to become too small and the network to lose its adaptivity (ability to move around) in the parameter space. Although regularization is typically understood from an overfitting perspective, we highlight its role in making the network more adaptive and enabling it to escape more easily from weights that generalize poorly. To provide such a capability, we propose a method called Logit Attenuating Weight Normalization (LAWN), that can be stacked onto any gradient-based optimizer. LAWN controls the logits by constraining the weight norms of layers in the final homogeneous sub-network. Empirically, we show that the resulting LAWN variant of the optimizer makes a deep network more adaptive to finding minimas with superior generalization performance on large-scale image classification and recommender systems. While LAWN is particularly impressive in improving Adam, it greatly improves all optimizers when used with large batch sizes

We present two large datasets of labelled plant-images that are suited towards the training of machine learning and computer vision models. The first dataset encompasses as the day of writing over 1.2 million images of indoor-grown crops and weeds common to the Canadian Prairies and many US states. The second dataset consists of over 540,000 images of plants imaged in farmland. All indoor plant images are labelled by species and we provide rich etadata on the level of individual images. This comprehensive database allows to filter the datasets under user-defined specifications such as for example the crop-type or the age of the plant. Furthermore, the indoor dataset contains images of plants taken from a wide variety of angles, including profile shots, top-down shots, and angled perspectives. The images taken from plants in fields are all from a top-down perspective and contain usually multiple plants per image. For these images metadata is also available. In this paper we describe both datasets' characteristics with respect to plant variety, plant age, and number of images. We further introduce an open-access sample of the indoor-dataset that contains 1,000 images of each species covered in our dataset. These, in total 14,000 images, had been selected, such that they form a representative sample with respect to plant age and ndividual plants per species. This sample serves as a quick entry point for new users to the dataset, allowing them to explore the data on a small scale and find the parameters of data most useful for their application without having to deal with hundreds of thousands of individual images.

We present a unified approach to (both finite and unrestricted) worst-case optimal entailment of (unions of) conjunctive queries (U)CQs in the wide class of "locally-forward" description logics. The main technique that we employ is a generalisation of Lutz's spoiler technique, originally developed for CQ entailment in ALCHQ. Our result closes numerous gaps present in the literature, most notably implying ExpTime-completeness of (U)CQ-querying for any superlogic of ALC contained in ALCHbregQ, and, as we believe, is abstract enough to be employed as a black-box in many new scenarios.

Lung X-ray images, if processed using statistical and computational methods, can distinguish pneumonia from COVID-19. The present work shows that it is possible to extract lung X-ray characteristics to improve the methods of examining and diagnosing patients with suspected COVID-19, distinguishing them from malaria, dengue, H1N1, tuberculosis, and Streptococcus pneumonia. More precisely, an intelligent computational model was developed to process lung X-ray images and classify whether the image is of a patient with COVID-19. The images were processed and extracted their characteristics. These characteristics were the input data for an unsupervised statistical learning method, PCA, and clustering, which identified specific attributes of X-ray images with Covid-19. The introduction of statistical models allowed a fast algorithm, which used the X-means clustering method associated with the Bayesian Information Criterion (CIB). The developed algorithm efficiently distinguished each pulmonary pathology from X-ray images. The method exhibited excellent sensitivity. The average recognition accuracy of COVID-19 was 0.93 and 0.051.

A number of deep learning based algorithms have been proposed to recover high-quality videos from low-quality compressed ones. Among them, some restore the missing details of each frame via exploring the spatiotemporal information of neighboring frames. However, these methods usually suffer from a narrow temporal scope, thus may miss some useful details from some frames outside the neighboring ones. In this paper, to boost artifact removal, on the one hand, we propose a Recursive Fusion (RF) module to model the temporal dependency within a long temporal range. Specifically, RF utilizes both the current reference frames and the preceding hidden state to conduct better spatiotemporal compensation. On the other hand, we design an efficient and effective Deformable Spatiotemporal Attention (DSTA) module such that the model can pay more effort on restoring the artifact-rich areas like the boundary area of a moving object. Extensive experiments show that our method outperforms the existing ones on the MFQE 2.0 dataset in terms of both fidelity and perceptual effect. Code is available at https://github.com/zhaominyiz/RFDA-PyTorch.

Each passing day, Museum and Cultural Places visitors' lives are subtly shaped by AI-driven technologies. In this smartphone glutted world, there lies a huge challenge for both the Museums and Cultural places to attract visitors. The question arises, What is the role of AI in a Museum? To learn more about visitors, manage visitor experience and collect relevant data for boosting the traffic and developing future growth strategies, Museums and Cultural Places across the globe are using artificial intelligence in several ways. The most commonly used modes of AI are - Robots & Chatboxes, Computer Visions and Natural language processing amongst others. Unlike traditional methods of managing generic data once a year, Museums rely on structured data to benefit both the visitors and the employees.

This will be a part of series of Machine Learning stories and this is the first one where we will cover few interesting but very basic concepts which is kind of must know for every budding data scientists or may be a professional one. A correlation coefficient tells you how strong, or how weak, the relationship is between two sets of data. In Mathematics, a coefficient is usually the number that is used to multiply a variable. So for this expression: 9x, the number 9 is the coefficient. A correlation between two variables or data sets indicates that as one variable changes in value, the other variable tends to change in a specific direction. It is also called the cross-correlation coefficient, Pearson correlation coefficient (PCC), or the Pearson product-moment correlation coefficient (PPMCC). Understanding this relationship is useful because the value of one variable allows us to predict the value of the other variable. For example, height and weight are correlated when it comes to your physique -- as height increases, the weight tends to increase too.

At first glance, a recently granted South African patent relating to a food container based on fractal geometry seems fairly mundane. The innovation in question involves interlocking food containers that are easy for robots to grasp and stack. On closer inspection, the patent is anything but mundane. That's because the inventor is not a human being -- it is an artificial intelligence (AI) system called DABUS. DABUS (which stands for device for the autonomous bootstrapping of unified sentience) is an AI system created by Stephen Thaler, a pioneer in the field of AI and programming.

Artificial Intelligence is, without a doubt, one of the Fourth Industrial Revolution's primary growth engines. The benefits and business potential inherent in this technology are immense. Improving customer experience, automating business processes, real-time information analysis, improving cyber protection capabilities, and implementing autonomous applications are just a few examples of these benefits. However, and similarly to other types of new and groundbreaking technologies, we must consider the latent risks in implementing Artificial Intelligence in an uncontrolled manner. Considering such risks is evidently even more urgent as Artificial Intelligence has now become so vastly used that it affects every aspect of our personal and professional life and used in scale, in large sectors of the economy.