Deaf-mute communities around the world experience a need in effective human-robot interaction system that would act as an interpreter in public places such as banks, hospitals, or police stations. The focus of this work is to address the challenges presented to hearing-impaired people by developing an interpreting robotic system required for effective communication in public places. To this end, we utilize a previously developed neural network-based learning architecture to recognize Cyrillic manual alphabet, which is used for finger spelling in Kazakhstan. In order to train and test the performance of the recognition system, we collected a depth data set of ten people and applied it to a learning-based method for gesture recognition by modeling motion data. We report our results that show an average accuracy of 77.2% for a complete alphabet recognition consisting of 33 letters.

Amid the wall-to-wall coverage of the U.S. presidential race, it was easy to miss the Obama administration's release this month of a slim, 48-page report titled "Preparing for the Future of Artificial Intelligence." Yet the subject of the report -- and the changes it foreshadows -- may prove to be as consequential for our society, and our education system, as even the most high-stakes national election. The term "artificial intelligence" means different things to different people, but broadly speaking, it refers to computers and advanced machines that can think, reason and communicate like humans, respond to novel or nuanced situations as a person might, and most critically, learn from experiences as a human would. According to a recent survey, 80 percent of AI researchers believe that computers and advanced machines will eventually achieve levels of artificial intelligence that rival human intelligence. Moreover, half believe that this will happen by the year 2040 -- just one generation from now.

As part of a three-week pilot study this quarter, half of the students in STATS 60: "Introduction to Statistical Methods" were assigned a RoboTA, an artificially intelligent robot teaching assistant (TA), to answer their questions by email. Students emailed different addresses based on the TA they were assigned. The emails were then funneled through a program, stripped of any identifying information and relayed to Lucas Janson, the TA for the class. Janson replied to all the emails without knowing whether they were intended for the human TA or the artificial intelligence (AI) TA. This study is double-blind, so neither the students nor the experimenters have information about the other, which helps prevent bias.

Intel announced a slew of products, technologies and investment in an effort to fix its position in the field of artificial intelligence. In the new move, Intel has assembled a set of technology options to drive AI capabilities in everything from smart factories and drones to sports, fraud detection and autonomous cars. Intel is increasing its focus on AI as it believes it can power the AI products released recently by companies like Facebook and Google. In a blog Intel CEO Brian Krzanich had said, "Intel is uniquely capable of enabling and accelerating the promise of AI. Intel is committed to AI and is making major investments in technology and developer resources to advance AI for business and society."

Artificial intelligence (AI) - machines that can learn from their results to improve their programming for better results. This is how humans operate. We try something, we judge the result and modify our behavior. What some considered to be science fiction only a few years ago, AI is edging closer to reality as decades of research, combined with advances in compute power, memory, storage, network connectivity, sensors, and the software that unites them all, is poised to enable new classes of intelligent predictive analytics. These innovations will bring benefits to multiple industries, and to society as a whole in the way we lead our everyday lives.

An exciting opportunity is available for a PhD candidate to undertake a research project in modelling autonomous behaviours, testing and verification of agent designs, explaining autonomous behaviour, or designing reusable simulation models. This scholarship is valued at up to $5000 per annum for up to 3 years. Students with an approved Australian Postgraduate Award (APA) or other postgraduate stipend are eligible for the top-up scholarship. Applicants should contact Associate Professor John Thangarajah to discuss their eligibility and the topic/area of prospective research (see further information below). These scholarships are most suited for those who plan to apply for an APA.

Saya had been teaching for seven years. Her impressive but short CV included stints in a few rural areas, overseas and as a substitute teacher. The difference is Saya is a remote controlled robot who taught her first class of 10-year olds in 2009. While we've all heard and read the stories of manual or labour-type jobs easily replaced by robots, increasingly the jobs we previously thought safe are no longer -- teachers, bankers, data analysts and the like are all at risk. But what do we really have to fear?

Diffuse interface methods have recently been introduced for the task of semi-supervised learning. The underlying model is well-known in materials science but was extended to graphs using a Ginzburg--Landau functional and the graph Laplacian. We here generalize the previously proposed model by a non-smooth potential function. Additionally, we show that the diffuse interface method can be used for the segmentation of data coming from hypergraphs. For this we show that the graph Laplacian in almost all cases is derived from hypergraph information. Additionally, we show that the formerly introduced hypergraph Laplacian coming from a relaxed optimization problem is well suited to be used within the diffuse interface method. We present computational experiments for graph and hypergraph Laplacians.

Classifiers for the semi-supervised setting often combine strong supervised models with additional learning objectives to make use of unlabeled data. This results in powerful though very complex models that are hard to train and that demand additional labels for optimal parameter tuning, which are often not given when labeled data is very sparse. We here study a minimalistic multi-layer generative neural network for semi-supervised learning in a form and setting as similar to standard discriminative networks as possible. Based on normalized Poisson mixtures, we derive compact and local learning and neural activation rules. Learning and inference in the network can be scaled using standard deep learning tools for parallelized GPU implementation. With the single objective of likelihood optimization, both labeled and unlabeled data are naturally incorporated into learning. Empirical evaluations on standard benchmarks show, that for datasets with few labels the derived minimalistic network improves on all classical deep learning approaches and is competitive with their recent variants without the need of additional labels for parameter tuning. Furthermore, we find that the studied network is the best performing monolithic ('non-hybrid') system for few labels, and that it can be applied in the limit of very few labels, where no other system has been reported to operate so far.

Developing machine learning solutions that give a lift from your existing prediction algorithms is not an easy task. They require a multitude of activities to get it right including cleaning up the data, setting up the infrastructure, testing & re-testing the model & finally deploying the algorithm. Here are five machine learning services that can help reduce the pain of deploying your machine learning solution. Based on Microsoft's Azure Cloud Platform, Azure Machine Learning offers a streamlined experience for all data scientist skill levels, from setting up with only a web browser, to using drag and drop gestures and simple data flow graphs to set up experiments. Machine Learning Studio features a library of time-saving sample experiments, R and Python packages and best-in-class algorithms from Microsoft businesses like Xbox and Bing.