Enterprises continued to accelerate the adoption of AI and machine learning to solve product and business challenges and improve revenues in 2021. Meanwhile, AI startups have experienced significant growth, roping in major investments to improve their product offerings and meet the growing demand for AI solutions across sectors. In fact, data from CB Insights Research shows that while the number of equity funding deals in the global AI space this year is just slightly less than the last (2,384 deals in 2021 versus 2,450 in 2020), the amount of capital invested has almost doubled to $68 billion. As we head into 2022, here's a quick look back at the milestones that shaped the AI space over the past 12 months. To start the year, OpenAI announced DALL-E, a multimodal AI system that generated images from text.

We base our work on the teleosemantic modelling of concepts as abilities implementing the distinct functions of recognition and classification. Accordingly, we model two types of concepts - substance concepts suited for object recognition exploiting visual properties, and classification concepts suited for classification of substance concepts exploiting linguistically grounded properties. The goal in this paper is to demonstrate that object recognition can be construed as classification via visual properties, as distinct from work in mainstream computer vision. Towards that, we present an object recognition process based on Ranganathan's four-phased faceted knowledge organization process, grounded in the teleosemantic distinctions of substance concept and classification concept. We also briefly introduce the ongoing project MultiMedia UKC, whose aim is to build an object recognition resource following our proposed process.

Modern neural language models widely used in tasks across NLP risk memorizing sensitive information from their training data. As models continue to scale up in parameters, training data, and compute, understanding memorization in language models is both important from a learning-theoretical point of view, and is practically crucial in real world applications. An open question in previous studies of memorization in language models is how to filter out "common" memorization. In fact, most memorization criteria strongly correlate with the number of occurrences in the training set, capturing "common" memorization such as familiar phrases, public knowledge or templated texts. In this paper, we provide a principled perspective inspired by a taxonomy of human memory in Psychology. From this perspective, we formulate a notion of counterfactual memorization, which characterizes how a model's predictions change if a particular document is omitted during training. We identify and study counterfactually-memorized training examples in standard text datasets. We further estimate the influence of each training example on the validation set and on generated texts, and show that this can provide direct evidence of the source of memorization at test time.

What AI and data analytics trends are taking the industry by storm this year? This comprehensive review highlights upcoming directions in AI to carefully watch and consider implementing in your personal work or organization.

The graph represents a network of 3,049 Twitter users whose tweets in the requested range contained "#cx", or who were replied to or mentioned in those tweets. The network was obtained from the NodeXL Graph Server on Monday, 13 December 2021 at 01:26 UTC. The requested start date was Sunday, 12 December 2021 at 01:01 UTC and the maximum number of days (going backward) was 14. The maximum number of tweets collected was 7,500. The tweets in the network were tweeted over the 3-day, 21-hour, 47-minute period from Wednesday, 08 December 2021 at 03:11 UTC to Sunday, 12 December 2021 at 00:59 UTC.

This paper presents a grounded language-image pre-training (GLIP) model for learning object-level, language-aware, and semantic-rich visual representations. GLIP unifies object detection and phrase grounding for pre-training. The unification brings two benefits: 1) it allows GLIP to learn from both detection and grounding data to improve both tasks and bootstrap a good grounding model; 2) GLIP can leverage massive image-text pairs by generating grounding boxes in a self-training fashion, making the learned representation semantic-rich. In our experiments, we pre-train GLIP on 27M grounding data, including 3M human-annotated and 24M web-crawled image-text pairs. The learned representations demonstrate strong zero-shot and few-shot transferability to various object-level recognition tasks. 1) When directly evaluated on COCO and LVIS (without seeing any images in COCO during pre-training), GLIP achieves 49.8 AP and 26.9 AP, respectively, surpassing many supervised baselines. 2) After fine-tuned on COCO, GLIP achieves 60.8 AP on val and 61.5 AP on test-dev, surpassing prior SoTA. 3) When transferred to 13 downstream object detection tasks, a 1-shot GLIP rivals with a fully-supervised Dynamic Head. Code will be released at https://github.com/microsoft/GLIP.

With the emergence of deep learning in the last years, new opportunities arose in Earth observation research. Nevertheless, they also brought with them new challenges. The data-hungry training processes of deep learning models demand large, resource expensive, annotated datasets and partly replaced knowledge-driven approaches, so that model behaviour and the final prediction process became a black box. The proposed SyntEO approach enables Earth observation researchers to automatically generate large deep learning ready datasets and thus free up otherwise occupied resources. SyntEO does this by including expert knowledge in the data generation process in a highly structured manner. In this way, fully controllable experiment environments are set up, which support insights in the model training. Thus, SyntEO makes the learning process approachable and model behaviour interpretable, an important cornerstone for explainable machine learning. We demonstrate the SyntEO approach by predicting offshore wind farms in Sentinel-1 images on two of the worlds largest offshore wind energy production sites. The largest generated dataset has 90,000 training examples. A basic convolutional neural network for object detection, that is only trained on this synthetic data, confidently detects offshore wind farms by minimising false detections in challenging environments. In addition, four sequential datasets are generated, demonstrating how the SyntEO approach can precisely define the dataset structure and influence the training process. SyntEO is thus a hybrid approach that creates an interface between expert knowledge and data-driven image analysis.

Considering the models that apply the contextual information of time-series data could improve the fault diagnosis performance, some neural network structures such as RNN, LSTM, and GRU were proposed to model the industrial process effectively. However, these models are restricted by their serial computation and hence cannot achieve high diagnostic efficiency. Also the parallel CNN is difficult to implement fault diagnosis in an efficient way because it requires larger convolution kernels or deep structure to achieve long-term feature extraction capabilities. Besides, BERT model applies absolute position embedding to introduce contextual information to the model, which would bring noise to the raw data and therefore cannot be applied to fault diagnosis directly. In order to address the above problems, a fault diagnosis model named deep parallel time-series relation network(\textit{DPTRN}) has been proposed in this paper. There are mainly three advantages for DPTRN: (1) Our proposed time relationship unit is based on full multilayer perceptron(\textit{MLP}) structure, therefore, DPTRN performs fault diagnosis in a parallel way and improves computing efficiency significantly. (2) By improving the absolute position embedding, our novel decoupling position embedding unit could be applied on the fault diagnosis directly and learn contextual information. (3) Our proposed DPTRN has obvious advantage in feature interpretability. Our model outperforms other methods on both TE and KDD-CUP99 datasets which confirms the effectiveness, efficiency and interpretability of the proposed DPTRN model.

Narrative cartography is a discipline which studies the interwoven nature of stories and maps. However, conventional geovisualization techniques of narratives often encounter several prominent challenges, including the data acquisition & integration challenge and the semantic challenge. To tackle these challenges, in this paper, we propose the idea of narrative cartography with knowledge graphs (KGs). Firstly, to tackle the data acquisition & integration challenge, we develop a set of KG-based GeoEnrichment toolboxes to allow users to search and retrieve relevant data from integrated cross-domain knowledge graphs for narrative mapping from within a GISystem. With the help of this tool, the retrieved data from KGs are directly materialized in a GIS format which is ready for spatial analysis and mapping. Two use cases - Magellan's expedition and World War II - are presented to show the effectiveness of this approach. In the meantime, several limitations are identified from this approach, such as data incompleteness, semantic incompatibility, and the semantic challenge in geovisualization. For the later two limitations, we propose a modular ontology for narrative cartography, which formalizes both the map content (Map Content Module) and the geovisualization process (Cartography Module). We demonstrate that, by representing both the map content and the geovisualization process in KGs (an ontology), we can realize both data reusability and map reproducibility for narrative cartography.

In recent years, the number of IoT devices has been growing fast which leads to a challenging task for managing, storing, analyzing, and making decisions about raw data from different IoT devices, especially for delay-sensitive applications. In a vehicular network (VANET) environment, the dynamic nature of vehicles makes the current open research issues even more challenging due to the frequent topology changes that can lead to disconnections between vehicles. To this end, a number of research works have been proposed in the context of cloud and fog computing over the 5G infrastructure. On the other hand, there are a variety of research proposals that aim to extend the connection time between vehicles. Vehicular Social Networks (VSNs) have been defined to decrease the burden of connection time between the vehicles. This survey paper first provides the necessary background information and definitions about fog, cloud and related paradigms such as 5G and SDN. Then, it introduces the reader to Vehicular Social Networks, the different metrics and the main differences between VSNs and Online Social Networks. Finally, this survey investigates the related works in the context of VANETs that have demonstrated different architectures to address the different issues in fog computing. Moreover, it provides a categorization of the different approaches and discusses the required metrics in the context of fog and cloud and compares them to Vehicular social networks. A comparison of the relevant related works is discussed along with new research challenges and trends in the domain of VSNs and fog computing.