All-natural and organic are familiar terms to consumers, and anything artificial has become anathema to many. Unless we're talking artificial intelligence – or AI – then investors should be hungry to learn as much as possible about a technology that is becoming as ubiquitous as organic tofu. The vast majority of nearly 2,000 experts polled by the Pew Research Center in 2014 said they anticipate robotics and artificial intelligence will permeate wide segments of daily life by 2025. A 2015 study covering 17 countries found that artificial intelligence and related technologies added an estimated 0.4 percentage point on average to those countries' annual GDP growth between 1993 and 2007, accounting for just over one-tenth of those countries' overall GDP growth during that time. Interesting numbers – but just what is artificial intelligence?

Quantum annealing (QA) is a hardware-based heuristic optimization and sampling method applicable to discrete undirected graphical models. While similar to simulated annealing, QA relies on quantum, rather than thermal, effects to explore complex search spaces. For many classes of problems, QA is known to offer computational advantages over simulated annealing. Here we report on the ability of recent QA hardware to accelerate training of fully visible Boltzmann machines. We characterize the sampling distribution of QA hardware, and show that in many cases, the quantum distributions differ significantly from classical Boltzmann distributions. In spite of this difference, training (which seeks to match data and model statistics) using standard classical gradient updates is still effective. We investigate the use of QA for seeding Markov chains as an alternative to contrastive divergence (CD) and persistent contrastive divergence (PCD). Using $k=50$ Gibbs steps, we show that for problems with high-energy barriers between modes, QA-based seeds can improve upon chains with CD and PCD initializations. For these hard problems, QA gradient estimates are more accurate, and allow for faster learning. Furthermore, and interestingly, even the case of raw QA samples (that is, $k=0$) achieved similar improvements. We argue that this relates to the fact that we are training a quantum rather than classical Boltzmann distribution in this case. The learned parameters give rise to hardware QA distributions closely approximating classical Boltzmann distributions that are hard to train with CD/PCD.

Secure Aggregation protocols allow a collection of mutually distrust parties, each holding a private value, to collaboratively compute the sum of those values without revealing the values themselves. We consider training a deep neural network in the Federated Learning model, using distributed stochastic gradient descent across user-held training data on mobile devices, wherein Secure Aggregation protects each user's model gradient. We design a novel, communication-efficient Secure Aggregation protocol for high-dimensional data that tolerates up to 1/3 users failing to complete the protocol. For 16-bit input values, our protocol offers 1.73x communication expansion for $2^{10}$ users and $2^{20}$-dimensional vectors, and 1.98x expansion for $2^{14}$ users and $2^{24}$ dimensional vectors.

In this paper, we deal with two challenges for measuring the similarity of the subject identities in practical video-based face recognition - the variation of the head pose in uncontrolled environments and the computational expense of processing videos. Since the frame-wise feature mean is unable to characterize the pose diversity among frames, we define and preserve the overall pose diversity and closeness in a video. Then, identity will be the only source of variation across videos since the pose varies even within a single video. Instead of simply using all the frames, we select those faces whose pose point is closest to the centroid of the K-means cluster containing that pose point. Then, we represent a video as a bag of frame-wise deep face features while the number of features has been reduced from hundreds to K. Since the video representation can well represent the identity, now we measure the subject similarity between two videos as the max correlation among all possible pairs in the two bags of features. On the official 5,000 video-pairs of the YouTube Face dataset for face verification, our algorithm achieves a comparable performance with VGG-face that averages over deep features of all frames. Other vision tasks can also benefit from the generic idea of employing geometric cues to improve the descriptiveness of deep features.

How can we efficiently propagate uncertainty in a latent state representation with recurrent neural networks? This paper introduces stochastic recurrent neural networks which glue a deterministic recurrent neural network and a state space model together to form a stochastic and sequential neural generative model. The clear separation of deterministic and stochastic layers allows a structured variational inference network to track the factorization of the model's posterior distribution. By retaining both the nonlinear recursive structure of a recurrent neural network and averaging over the uncertainty in a latent path, like a state space model, we improve the state of the art results on the Blizzard and TIMIT speech modeling data sets by a large margin, while achieving comparable performances to competing methods on polyphonic music modeling.

Generative Adversarial Networks (GAN) have limitations when the goal is to generate sequences of discrete elements. The reason for this is that samples from a distribution on discrete objects such as the multinomial are not differentiable with respect to the distribution parameters. This problem can be avoided by using the Gumbel-softmax distribution, which is a continuous approximation to a multinomial distribution parameterized in terms of the softmax function. In this work, we evaluate the performance of GANs based on recurrent neural networks with Gumbel-softmax output distributions in the task of generating sequences of discrete elements.

Lip reading is a way of understanding speech by interpreting a person's lip movement. However, human speech is highly complex and nuanced, where one lip movement could correspond to different phonemes, or basic units of sound. Therefore, the practice is prone to errors, which can sometimes lead to humorous results. Scientists from Oxford University have described an artificial intelligence system, called LipNet, which can accurately read lips. The system employs deep learning to train itself using 29,000 three-second-long videos labeled with captions.

Artificial Intelligence is affecting almost every industry and is transforming the way businesses operate. The combination of new algorithms, big data, and GPUs has made it possible to address problems that were not practically solvable until now. During this webinar we'll provide an overview of the different AI and deep learning applications for UAVs, including warehouse management, aerial inspection, search and rescue, and agriculture, and explain how these applications can be easily deployed via Jetson.

Under, behind and inside many of the apps we use every day, a revolution is underway. It's a revolution that started decades ago but today is empowering companies to deliver better, smarter services with greater ease and on broader scales than ever before. At Singularity University's inaugural Global Summit, Neil Jacobstein, chair of Artificial Intelligence and Robotics, provided a primer showing how artificial intelligence literally transforms everything it touches. First of all, it's critical to define the scope of artificial intelligence (AI), which can be categorized into four areas: techniques in pattern recognition, software agency (that is, software that acts like real users), an exponential technology that is accelerating other exponential technologies, and a vision of a future superhuman intelligence (that fortunately hasn't happened yet). Anyone who has seen a science fiction film is likely familiar with this last area, but it's the other three areas where AI is making huge strides at a revolutionary pace.

A BSTRACT Restricted Boltzmann Machine (RBM) is a bipartite graphical model that is used as the building block in energy-based deep generative models. Due to numerical stability and quantifiability of the likelihood, RBM is commonly used with Bernoulli units. Here, we consider an alternative member of exponential family RBM with leaky rectified linear units - called leaky RBM. We first study the joint and marginal distributions of leaky RBM under different leakiness, which provides us important insights by connecting the leaky RBM model and truncated Gaussian distributions. The connection leads us to a simple yet efficient method for sampling from this model, where the basic idea is to anneal the leakiness rather than the energy; - i.e., start from a fully Gaussian/Linear unit and gradually decrease the leakiness over iterations. This serves as an alternative to the annealing of the temperature parameter and enables numerical estimation of the likelihood that are more efficient and more accurate than the commonly used annealed importance sampling (AIS). We further demonstrate that the proposed sampling algorithm enjoys faster mixing property than contrastive divergence algorithm, which benefits the training without any additional computational cost. 1 I NTRODUCTION In this paper, we are interested in deep generative models. There is a family of directed deep generative models which can be trained by back-propagation (e.g., Kingma & Welling, 2013; Goodfellow et al., 2014). The other family is the deep energy-based models, including deep belief network (Hinton et al., 2006) and deep Boltzmann machine (Salakhutdinov & Hinton, 2009).