The release of two machine learning (ML) model builders have made it easier for software engineers to create and run ML models, even without specialized training. Microsoft and Amazon Web Services' (AWS) Gluon is an open source project that eliminates some of the difficult work required to develop artificial intelligence (AI) systems. It provides training algorithms and neural network models, two important components of a deep learning system, that developers can use to develop their own ML systems. Google's ML engine is part of its cloud platform and is offered as a managed service for developers to build ML models that work on any type of data, of any size. Similar to Gluon, Google's service provides pre-trained models for developers to generate their own tailored ML models.

Deep learning has revolutionized how we process the vast firehoses of data that define modern life. Yet, the daily drumbeat of AI headlines tends to center on the commercial applications of AI and how it is reshaping how companies do business. In a refreshing twist, a new open AI challenge by the World Bank, in collaboration with WeRobotics and OpenAerialMap, illustrates the incredible potential of deep learning for humanitarian applications, especially in the critical hours and days after a major natural disaster. One of the most exciting application areas of modern deep learning tools has been the use of neural networks to examine imagery at accuracy and detail levels impossible just a few years ago. Today state-of-the-art neural systems can examine hundreds of millions of images, cataloging them into tens of thousands of categories, estimating the location they were taken, their emotion, look in the background for pollution and natural disaster damage and even estimate the level of "violence" they portray, while creating new models is increasingly becoming point-and-click.

In August 2015, a number of carefully selected Facebook users in the Bay Area discovered a new feature on Facebook Messenger. Known as M, the service was designed to rival Google Now and Apple's Siri. A personal assistant that would answer questions in a natural way, make restaurant reservations and help with Uber bookings, M was meant to be a step forward in natural language understanding, the virtual assistant that – unlike Siri – wasn't a dismal experience. Fast forward a couple of years, and the general purpose personal assistant has been demoted within Facebook's product offering. Poor M. The hope was that it would tell users jokes and act as a guide, life coach and optimisation tool.

WATERxRIVAL is the confluence my current investigations (collaboration with a deep learning AI trained on my artwork and image archive, named RIVAL) and a previous investigation (speculative web-page about water evaporation as a 4th-dimensional phenomenon). This convergence becomes an exploration more ecological in scope and perspective. RIVAL's algorithmic flows/processes are let out into the open waters of internet image & keyword searches, returning with specimens and samples for me to compose and cultivate meaning(fulness) with. The initial set of images (from the speculative exploration of water and space) are put into relationship with the ocean of media flooding all around us. The WATERxRIVAL webpage provides an opportunity for the viewer to glimpse the deluge (this process produces); additionally, it will expand and swell weekly throughout the exhibition's duration.

The IEEE publishes an annual list of the Top 10 Technology Trends for each upcoming year. Making the list for 2018 are multiple topics surrounding artificial intelligence and machine learning. Deep learning comes in as the IEEE hottest trend for 2018. Neural networks extract features through a concept of layers. By combining the output from these multiple layers, deeper layers are able to construct more advanced insight from data.

We are interested in the development of surrogate models for uncertainty quantification and propagation in problems governed by stochastic PDEs using a deep convolutional encoder-decoder network in a similar fashion to approaches considered in deep learning for image-to-image regression tasks. Since normal neural networks are data intensive and cannot provide predictive uncertainty, we propose a Bayesian approach to convolutional neural nets. A recently introduced variational gradient descent algorithm based on Stein's method is scaled to deep convolutional networks to perform approximate Bayesian inference on millions of uncertain network parameters. This approach achieves state of the art performance in terms of predictive accuracy and uncertainty quantification in comparison to other approaches in Bayesian neural networks as well as techniques that include Gaussian processes and ensemble methods even when the training data size is relatively small. To evaluate the performance of this approach, we consider standard uncertainty quantification benchmark problems including flow in heterogeneous media defined in terms of limited data-driven permeability realizations. The performance of the surrogate model developed is very good even though there is no underlying structure shared between the input (permeability) and output (flow/pressure) fields as is often the case in the image-to-image regression models used in computer vision problems. Studies are performed with an underlying stochastic input dimensionality up to $4,225$ where most other uncertainty quantification methods fail. Uncertainty propagation tasks are considered and the predictive output Bayesian statistics are compared to those obtained with Monte Carlo estimates.

Deep learning has recently seen rapid development and significant attention due to its state-of-the-art performance on previously-thought hard problems. However, because of the innate complexity and nonlinear structure of deep neural networks, the underlying decision making processes for why these models are achieving such high performance are challenging and sometimes mystifying to interpret. As deep learning spreads across domains, it is of paramount importance that we equip users of deep learning with tools for understanding when a model works correctly, when it fails, and ultimately how to improve its performance. Standardized toolkits for building neural networks have helped democratize deep learning; visual analytics systems have now been developed to support model explanation, interpretation, debugging, and improvement. We present a survey of the role of visual analytics in deep learning research, noting its short yet impactful history and summarize the state-of-the-art using a human-centered interrogative framework, focusing on the Five W's and How (Why, Who, What, How, When, and Where), to thoroughly summarize deep learning visual analytics research. We conclude by highlighting research directions and open research problems. This survey helps new researchers and practitioners in both visual analytics and deep learning to quickly learn key aspects of this young and rapidly growing body of research, whose impact spans a diverse range of domains.

Scene recognition with RGB images has been extensively studied and has reached very remarkable recognition levels, thanks to convolutional neural networks (CNN) and large scene datasets. In contrast, current RGB-D scene data is much more limited, so often leverages RGB large datasets, by transferring pretrained RGB CNN models and fine-tuning with the target RGB-D dataset. However, we show that this approach has the limitation of hardly reaching bottom layers, which is key to learn modality-specific features. In contrast, we focus on the bottom layers, and propose an alternative strategy to learn depth features combining local weakly supervised training from patches followed by global fine tuning with images. This strategy is capable of learning very discriminative depth-specific features with limited depth images, without resorting to Places-CNN. In addition we propose a modified CNN architecture to further match the complexity of the model and the amount of data available. For RGB-D scene recognition, depth and RGB features are combined by projecting them in a common space and further leaning a multilayer classifier, which is jointly optimized in an end-to-end network. Our framework achieves state-of-the-art accuracy on NYU2 and SUN RGB-D in both depth only and combined RGB-D data.

From hurricanes to wildfires, 2017 brought the world a number of natural disasters -- as well as some tech to deal with them. We have more information than ever following a disaster thanks to unmanned aerial vehicles (UAVs) and sophisticated satellites that can capture images of disasters from the air, but we are still working on ways to process the data so it is valuable for relief efforts. That's where deep learning comes in, says the World Bank in collaboration with WeRobotics and OpenAerialMap. On Jan. 10, 2018, World Bank issued an artificial intelligence (AI) challenge to explore how deep learning could be used in the wake of natural disasters. Deep learning is what enables AI to recognize patterns in images, sounds, and other data using a neural network that mirrors our own grey matter.

Abstract: "Visual recognition involves reasoning about structured relations at multiple levels of detail. For example, human behaviour analysis requires a comprehensive labeling covering individual low-level actions to pair-wise interactions through to high-level events. Scene understanding can benefit from considering labels and their inter-relations. In this talk I will present recent work by our group building deep learning approaches capable of modeling these structures. I will present models for learning trajectory features that represent individual human actions, and hierarchical temporal models for group activity recognition. General purpose structured inference machines will be described, building from notions of message passing within graphical models. These will be used in models for inferring individual and group activity and modeling structured relations for image labeling problems."