Bushfires in Australia are as commonplace as kangaroos and koalas. A hot, dry climate regularly sets the stage for conflagrations that endanger human lives, property, and wildlife and threaten one of the country's top economic industries: wine. Fires during summer 2019–2020 decimated entire vineyards in South Australia, Victoria and New South Wales, but smoke, which was far more widespread and insidious, seeped into grapes and into fermenting barrels, yielding unpleasant, unsaleable product. Although the full extent of the damage caused has not yet been calculated, analysis from the Australian Wine Research Institute indicates that smoke taint alone costs the country's wine industry tens to hundreds of millions of dollars each time a high fire season occurs. Advances in a wide range of technologies could help growers and winemakers mitigate the negative impact of smoke taint and other unpredictable anomalies, such as frost, drought, pests, and disease -- and not just in Australia, but around the world.

The speech-to-text technology has made our lives easy. Now that there are a lot of use cases in our daily life for this technology too, we need it even more. It lets us save time and effort, and provides the required information in a matter of minutes. Tech giants like Google and Amazon are exploring and empowering this field of speech recognition technologies with the help of their Google Speech and Amazon Transcribe products. Amazon launched Alexa in 2014 and more than 100m of its Echo and Dot gadgets are available in homes around the world today. Alexa is considered to be the most Intelligent Of All DPAs.

Facebook says it has deployed a feature in its Community Help hub to make it easier for users to assist each other during the pandemic. As of this week, AI will detect when a public post on News Feed is about needing or offering help and will surface a suggestion to share it on Community Help. Once a post is moved or published directly to the hub, an algorithm will recommend matches between people. For example, if someone posts an offer to deliver groceries, they'll see recommendations within Community Help to connect with people who recently posted about needing this type of assistance. Similarly, if someone requests masks, AI will surface suggested neighbors who recently posted an offer to make face coverings.

The field of machine translation (MT), the automatic translation of written text from one natural language into another, has experienced a major paradigm shift in recent years. Statistical MT, which mainly relies on various count-based models and which used to dominate MT research for decades, has largely been superseded by neural machine translation (NMT), which tackles translation with a single neural network. In this work we will trace back the origins of modern NMT architectures to word and sentence embeddings and earlier examples of the encoder-decoder network family. We will conclude with a short survey of more recent trends in the field.

While unsupervised domain translation (UDT) has seen a lot of success recently, we argue that allowing its translation to be mediated via categorical semantic features could enable wider applicability. In particular, we argue that categorical semantics are important when translating between domains with multiple object categories possessing distinctive styles, or even between domains that are simply too different but still share high-level semantics. We propose a method to learn, in an unsupervised manner, categorical semantic features (such as object labels) that are invariant of the source and target domains. We show that conditioning the style of a unsupervised domain translation methods on the learned categorical semantics leads to a considerably better high-level features preservation on tasks such as MNIST$\leftrightarrow$SVHN and to a more realistic stylization on Sketches$\to$Reals.

Entities are at the center of how we represent and aggregate knowledge. For instance, Encyclopedias such as Wikipedia are structured by entities (e.g., one per article). The ability to retrieve such entities given a query is fundamental for knowledge-intensive tasks such as entity linking and open-domain question answering. One way to understand current approaches is as classifiers among atomic labels, one for each entity. Their weight vectors are dense entity representations produced by encoding entity information such as descriptions. This approach leads to several shortcomings: i) context and entity affinity is mainly captured through a vector dot product, potentially missing fine-grained interactions between the two; ii) a large memory footprint is needed to store dense representations when considering large entity sets; iii) an appropriately hard set of negative data has to be subsampled at training time. We propose GENRE, the first system that retrieves entities by generating their unique names, left to right, token-by-token in an autoregressive fashion, and conditioned on the context. This enables to mitigate the aforementioned technical issues: i) the autoregressive formulation allows us to directly capture relations between context and entity name, effectively cross encoding both; ii) the memory footprint is greatly reduced because the parameters of our encoder-decoder architecture scale with vocabulary size, not entity count; iii) the exact softmax loss can be efficiently computed without the need to subsample negative data. We show the efficacy of the approach with more than 20 datasets on entity disambiguation, end-to-end entity linking and document retrieval tasks, achieving new SOTA, or very competitive results while using a tiny fraction of the memory of competing systems. Finally, we demonstrate that new entities can be added by simply specifying their unambiguous name.

This work demonstrates how a complete statistical model quantifying the evolution of pairs of aligned proteins can be constructed from a time-parameterised substitution matrix and a time-parameterised 3-state alignment machine. All parameters of such a model can be inferred from any benchmark data-set of aligned protein sequences. This allows us to examine nine well-known substitution matrices on six benchmarks curated using various structural alignment methods; any matrix that does not explicitly model a "time"-dependent Markov process is converted to a corresponding base-matrix that does. In addition, a new optimal matrix is inferred for each of the six benchmarks. Using Minimum Message Length (MML) inference, all 15 matrices are compared in terms of measuring the Shannon information content of each benchmark. This has resulted in a new and clear overall best performed time-dependent Markov matrix, MMLSUM, and its associated 3-state machine, whose properties we have analysed in this work. For standard use, the MMLSUM series of (log-odds) \textit{scoring} matrices derived from the above Markov matrix, are available at https://lcb.infotech.monash.edu.au/mmlsum.

Learning controllable and generalizable representation of multivariate data with desired structural properties remains a fundamental problem in machine learning. In this paper, we present a novel framework for learning generative models with various underlying structures in the latent space. We represent the inductive bias in the form of mask variables to model the dependency structure in the graphical model and extend the theory of multivariate information bottleneck to enforce it. Our model provides a principled approach to learn a set of semantically meaningful latent factors that reflect various types of desired structures like capturing correlation or encoding invariance, while also offering the flexibility to automatically estimate the dependency structure from data. We show that our framework unifies many existing generative models and can be applied to a variety of tasks including multi-modal data modeling, algorithmic fairness, and invariant risk minimization.

An expanding field of substantive interest for the theory of the law and the practice-of-law entails Legal Sentiment Analysis and Opinion Mining (LSAOM), consisting of two often intertwined phenomena and actions underlying legal discussions and narratives: (1) Sentiment Analysis (SA) for the detection of expressed or implied sentiment about a legal matter within the context of a legal milieu, and (2) Opinion Mining (OM) for the identification and illumination of explicit or implicit opinion accompaniments immersed within legal discourse. Efforts to undertake LSAOM have historically been performed by human hand and cognition, and only thinly aided in more recent times by the use of computer-based approaches. Advances in Artificial Intelligence (AI) involving especially Natural Language Processing (NLP) and Machine Learning (ML) are increasingly bolstering how automation can systematically perform either or both of Sentiment Analysis and Opinion Mining, all of which is being inexorably carried over into engagement within a legal context for improving LSAOM capabilities. This research paper examines the evolving infusion of AI into Legal Sentiment Analysis and Opinion Mining and proposes an alignment with the Levels of Autonomy (LoA) of AI Legal Reasoning (AILR), plus provides additional insights regarding AI LSAOM in its mechanizations and potential impact to the study of law and the practicing of law.

Recent work in open-domain conversational agents has demonstrated that significant improvements in model engagingness and humanness metrics can be achieved via massive scaling in both pre-training data and model size (Adiwardana et al., 2020; Roller et al., 2020). However, if we want to build agents with human-like abilities, we must expand beyond handling just text. A particularly important topic is the ability to see images and communicate about what is perceived. With the goal of engaging humans in multi-modal dialogue, we investigate combining components from state-of-the-art open-domain dialogue agents with those from state-of-the-art vision models. We study incorporating different image fusion schemes and domain-adaptive pre-training and fine-tuning strategies, and show that our best resulting model outperforms strong existing models in multi-modal dialogue while simultaneously performing as well as its predecessor (text-only) BlenderBot (Roller et al., 2020) in text-based conversation. We additionally investigate and incorporate safety components in our final model, and show that such efforts do not diminish model performance with respect to engagingness metrics.