Time-resolved angiography with interleaved stochastic trajectories (TWIST) has been widely used for dynamic contrast enhanced MRI (DCE-MRI). To achieve highly accelerated acquisitions, TWIST combines the periphery of the k-space data from several adjacent frames to reconstruct one temporal frame. However, this view-sharing scheme limits the true temporal resolution of TWIST. In addition, since the k-space sampling patterns have been specially designed for a specific generalized autocalibrating partial parallel acquisition (GRAPPA) factor, it is not possible to reduce the number of views in order to reconstruct images with a better temporal resolution. To address these issues, this paper proposes a novel k-space deep learning approach for parallel MRI. In particular, inspired by the recent mathematical discovery that links Hankel matrix decomposition to deep learning, we have implemented our neural network so that accurate k-space interpolations are performed simultaneously for multiple coils by exploiting the redundancies along the coils and images. In addition, the proposed method can immediately generate reconstruction results with different numbers of view-sharing, allowing us to exploit the trade-off between spatial and temporal resolution. Reconstruction results using in vivo TWIST data set confirm the accuracy and the flexibility of the proposed method.

In recent years, there has been a surge of interest in developing deep learning methods for non-Euclidean structured data such as graphs. In this paper, we propose Dual-Primal Graph CNN, a graph convolutional architecture that alternates convolution-like operations on the graph and its dual. Our approach allows to learn both vertex- and edge features and generalizes the previous graph attention (GAT) model. We provide extensive experimental validation showing state-of-the-art results on a variety of tasks tested on established graph benchmarks, including CORA and Citeseer citation networks as well as MovieLens, Flixter, Douban and Yahoo Music graph-guided recommender systems.

We provide a new understanding of the fundamental nature of adversarially robust classifiers and how they differ from standard models. In particular, we show that there provably exists a trade-off between the standard accuracy of a model and its robustness to adversarial perturbations. We demonstrate an intriguing phenomenon at the root of this tension: a certain dichotomy between "robust" and "non-robust" features. We show that while robustness comes at a price, it also has some surprising benefits. Robust models turn out to have interpretable gradients and feature representations that align unusually well with salient data characteristics. In fact, they yield striking feature interpolations that have thus far been possible to obtain only using generative models such as GANs.

Neural attention (NA) is an effective mechanism for inferring complex structural data dependencies that span long temporal horizons. As a consequence, it has become a key component of sequence-to-sequence models that yield state-of-the-art performance in as hard tasks as abstractive document summarization (ADS), machine translation (MT), and video captioning (VC). NA mechanisms perform inference of context vectors; these constitute weighted sums of deterministic input sequence encodings, adaptively sourced over long temporal horizons. However, recent work in the field of amortized variational inference (AVI) has shown that it is often useful to treat the representations generated by deep networks as latent random variables. This allows for the models to better explore the space of possible representations. Based on this motivation, in this work we introduce a novel regard towards a popular NA mechanism, namely soft-attention (SA). Our approach treats the context vectors generated by SA models as latent variables, the posteriors of which are inferred by employing AVI. Both the means and the covariance matrices of the inferred posteriors are parameterized via deep network mechanisms similar to those employed in the context of standard SA. To illustrate our method, we implement it in the context of popular sequence-to-sequence model variants with SA. We conduct an extensive experimental evaluation using challenging ADS, VC, and MT benchmarks, and show how our approach compares to the baselines.

According to a recent paper in Neuron, our brains are meant to act as information filters. Put in a big pile of messy data, filter for the useful bits, then clear out any irrelevant details in order to tell a story or make a decision. The unused pieces are deleted in order to make space for new data -- like running a disk cleanup on a computer. In neurobiology terms, forgetting happens when synaptic connections between neurons weaken or are eliminated over time, and as new neurons develop, they rewire the circuits of the hippocampus, overwriting existing memories (New Atlas). In order to adapt effectively, humans need to be able to strategically forget.

Imec said at its annual event here that it is prototyping a deep-learning inference chip using single-bit precision. The research institute hopes to gather data over the next year on the effectiveness for client devices of the novel data type and architecture--either a processor-in-memory (PIM) or an analog memory fabric. The PIM architecture, explored by academics for decades, is gaining popularity for data-intensive machine-learning algorithms. Startup Mythic and IBM Research are designing two of the most prominent efforts in the field. Many academics are experimenting with 1- to 4-bit data types to trim the heavy memory requirements for deep learning.

One of the highlights of last year's Google I/O was the announcement of Google's Automated Machine Learning or AutoML which according to founder Sundar Pichai was their'AI Inception'. The Machine learning Cloud software suite is finally about to hit the alpha stage and as of now, it seems like there is a lot that developer and designers alike can leverage from this tool. So let us explore this new addition to the Google family. AutoML is Google's Cloud software suite for Machine Learning tools based on Google's Neural Architecture Search (NAS). Using AutoML a user can train deep networks without having any expertise in deep learning or Artificial Intelligence.

The biggest risk in world of cryptocurrencies is not taking any risk… In the ever-changing crypto markets, leverage trading was built for those who believe the only strategy guaranteed to fail is the one that entails no risks. We strongly recommend that you set your leverage at 50x and let our AI autonomously trade on your behalf. It harnesses the power of machine learning to read and interpret real-time patterns and makes swift trading decisions. In-built deep learning tools allow these decisions to be made on basis of all big data that we have collated for the AI to trade.

Coinvision sat down with Ben Goertzel, CEO and Chief Scientist of SingularityNET, a project leading the way in integrating the blockchain and artificial intelligence (AI) by creating a decentralized marketplace for knowledge sharing in AI research. After a USD$36 million initial coin offering that sold out in just over a minute, a lot of expectation surrounds one of the most forward looking projects in the blockchain sphere today. Yesterday was a big day for SingularityNET as the team is in Toronto to announce the launch a new platform named DAIA, an Industry Alliance Leveraging Blockchain to Democratize AI. The project comes from a partnership between SingularityNET and AIDecentralized and will be working with over 100 projects in AI development. This is what Ben had to stay about the state of affairs of SingularityNET right now.

Christopher Bouton, PhD, a self-described molecular neurobiologist turned entrepreneur, had such a positive experience starting and running a company that he decided to do it twice. After a five-year stint at pharmaceutical/biotech giant, Pfizer, the Johns Hopkins grad started a company called Entagen, which developed semantic-based analytics for the healthcare sector. After five years running Entagen, Thomson Reuters acquired the company in what Bouton called a "successful exit." Bouton spent a few years under the Thomson Reuters umbrella, but the entrepreneurial itch soon returned. "[In] 2016 while I was sort of contemplating what to do next, I started to take note of all of these deep learning approaches that were starting to be talked about. The reason that they're talking about AI is because of these deep learning algorithms. And so I got interested in what they were and how they worked."