Autonomous vehicles are poised to take to our roads in massive numbers in the coming years. This has been made possible due to advances in deep learning and its application to autonomous driving. In this post, we take you through a tutorial that shows you how to build a remote control (RC) vehicle that uses Amazon AI services. Typically each autonomous vehicle is stacked with a lot of sensors that provide rich telemetry. This telemetry can be used to improve the driving of the individual vehicle but also the user experience.

Crystallography is the researcher's tool of choice to reveal protein structure – but it can be slow and laborious. Crystallography has long been an indispensable tool for elucidating the structure of a wide variety of molecules. This is especially true of biological macromolecules, which has granted generations of researchers insight into how these large molecules work and interact. For pharmaceutical research, understanding the structure of a protein assists researchers in the rational design of new small and large molecule therapeutics as they can tailor their candidates to more effectively interact with their target. However, while the number of complete structures in protein structure databases has grown exponentially over the past decade, there still are gaps in these databases due to the laborious nature of determining a de novo protein structure.

As the wave of digital transformation and deep learning technologies continues to wash over the world of business, we are seeing significant changes in how companies are communicating with their customers. With a recent report claiming that the AI software market will grow from $1.38 billion in 2016 to $59.75 billion by 2025, it's clear that chatbots, machine learning and deep learning technologies will play a crucial role in the future of enterprise communications – but what exactly is it and what kind of impact will it have as the years go by? In order to answer these questions, we decided to write a short yet informative piece on deep learning technology and what it brings to the table for businesses in post-digital world. AI is not an unfamiliar concept and the idea of machines that can think has existed for as far back as the 1950s. However, over the last decade or so, we have seen a monumental shift in the development of new AI technologies that are changing the way people communicate and taking Artificial Intelligence to a whole new level. If we were to look at machine learning, a term used to describe the way a computer uses algorithms to analyse data, learn from it and then act on what it has learnt, we can see that AI works not by following a set of strict instructions set by a programmer, but instead functions independently to analyse wide sources of information available through the cloud and solve problems by itself.

"I believe that by the end of the century, the use of words will have been altered so much that one will be able to speak of machines thinking without expecting to be contradicted." The foundation of AI started back in 1950 when scientist Alan Turing published a seminal paper containing a description of what is now referred to as the "Turing Test" which is designed to determine if a machine can think. A group of scientists got together at Dartmouth College a few years later and coined the term "artificial intelligence". Fast forward to this year and a panel discussion at the Invest 2017 Conference held in New York City on the ability of AI to revolutionize wealth management. Implementing AI technology is an offensive move all the way, Clinc's Mars stated.

For the past several years, our goal at Google has been to play a critical role in bringing the benefits of AI to everyone. Machine learning is at the heart of that goal. In the area of life sciences -- or more specifically, the field of genomics -- we're using ML to derive insights from the human genome. Additionally, due to the scale of human genomic data, we need new techniques to process the datasets using machine learning and cloud computing. We'll examine an analytical technique for verifying the ancestry of a human DNA sample, and show you how to implement it as a system using Google Cloud Platform, TensorFlow and Google Cloud Machine Learning Engine.

The field of AI has made great strides in the last several years, thanks to developments in machine learning algorithms and deep learning systems based on artificial neural networks (ANNs). Researchers have found that vast sets of example data are the way to train up such systems to produce the desired results, whether that is picking out a face from a photograph or recognizing speech input. But the resultant systems often turn out to operate as an inscrutable "black box" and even their developers find themselves unable to explain why it arrived at a particular decision. That may soon prove unacceptable in areas where an AI's decisions could have an impact on people's lives, such as employment, mortgage lending, or self-driving vehicles. The value of so-called explainable AI was called into question recently by Google research director Peter Norvig, who noted that humans are not very good at explaining their decision-making either, and claimed that the performance of an AI system could be gauged simply by observing its outputs over time.

When classifying point clouds, a large amount of time is devoted to the process of engineering a reliable set of features which are then passed to a classifier of choice. Generally, such features - usually derived from the 3D-covariance matrix - are computed using the surrounding neighborhood of points. While these features capture local information, the process is usually time-consuming, and requires the application at multiple scales combined with contextual methods in order to adequately describe the diversity of objects within a scene. In this paper we present a 1D-fully convolutional network that consumes terrain-normalized points directly with the corresponding spectral data,if available, to generate point-wise labeling while implicitly learning contextual features in an end-to-end fashion. Our method uses only the 3D-coordinates and three corresponding spectral features for each point. Spectral features may either be extracted from 2D-georeferenced images, as shown here for Light Detection and Ranging (LiDAR) point clouds, or extracted directly for passive-derived point clouds,i.e. from muliple-view imagery. We train our network by splitting the data into square regions, and use a pooling layer that respects the permutation-invariance of the input points. Evaluated using the ISPRS 3D Semantic Labeling Contest, our method scored second place with an overall accuracy of 81.6%. We ranked third place with a mean F1-score of 63.32%, surpassing the F1-score of the method with highest accuracy by 1.69%. In addition to labeling 3D-point clouds, we also show that our method can be easily extended to 2D-semantic segmentation tasks, with promising initial results.

We consider the task of detecting regulatory elements in the human genome directly from raw DNA. Past work has focused on small snippets of DNA, making it difficult to model long-distance dependencies that arise from DNA's 3-dimensional conformation. In order to study long-distance dependencies, we develop and release a novel dataset for a larger-context modeling task. Using this new data set we model long-distance interactions using dilated convolutional neural networks, and compare them to standard convolutions and recurrent neural networks. We show that dilated convolutions are effective at modeling the locations of regulatory markers in the human genome, such as transcription factor binding sites, histone modifications, and DNAse hypersensitivity sites.

Deep learning models are impactful in many real-world applications, and the transfer of this technology to society has created new emerging issues, like the need of model interpretability [14]. The General Data Protection Regulation approved in 2016 by the European parliament, which will be effective in 2018, is a concrete example of the need to provide human understandable justifications for decisions taken by automated data-processing systems [15]. Research could probably be inspired by old literature in neural networks to find better explanations about the dynamics of deep learning and provide more human interpretable solutions. An example of such process is found in standard logistic activation functions, that have been studied extensively in the past, but tend to be substituted by other activation functions in modern neural networks. To understand why this may be the case, it is important to recall the unique properties of the logistic function and therefore analyze the reasons why it has been introduced in neural networks. Firstly, the standard logistic function is biologically plausible.

Facial Key Points (FKPs) detection is an important and challenging problem in the field of computer vision, which involves detecting FKPs like centers and corners of eyes, nose tip, etc. The problem is to predict the (x, y) realvalued coordinates in the space of image pixels of the FKPs for a given face image. It finds its application in tracking faces in images and videos, analysis of facial expressions, detection of dysmorphic facial signs for medical diagnosis, face recognition, etc. Facial features vary greatly from one individual to another, and even for a single individual there is a large amount of variation due to pose, size, position, etc. The problem becomes even more challenging when the face images are taken under different illumination conditions, viewing angles, etc. In the past few years, advancements in FKPs detection are made by the application of deep convolutional neural network (DCNN), which is a special type of feed-forward neural network with shared weights and local connectivity. DCNNs have helped build state-of-the-art models for image recognition, recommender systems, natural language processing, etc. Krizhevsky et al. [1] applied DCNN in ImageNet image classification challenge and outperformed the previous state-of-the-art model for image classification. Wang et al. [2] addressed FKPs detection by first applying histogram stretching for image contrast enhancement, followed by principal component analysis for noise reduction and mean patch search algorithm with correlation scoring and mutual information scoring for predicting left and right eye centers. Sun et al. [3] estimated FKPs by using a three level convolutional neural network, where at each level, outputs of multiple networks were fused for robust and accurate estimation. Longpre et al. [4] predicted FKPs by first applying data augmentation to expand the number of training examples, followed by testing different architectures of convolutional