Owing to their connection with generative adversarial networks (GANs), saddle-point problems have recently attracted considerable interest in machine learning and beyond. By necessity, most theoretical guarantees revolve around convex-concave problems; however, making theoretical inroads towards efficient GAN training crucially depends on moving beyond this classic framework. To make piecemeal progress along these lines, we analyze the widely used mirror descent (MD) method in a class of non-monotone problems - called coherent - whose solutions coincide with those of a naturally associated variational inequality. Our first result is that, under strict coherence (a condition satisfied by all strictly convex-concave problems), MD methods converge globally; however, they may fail to converge even in simple, bilinear models. To mitigate this deficiency, we add on an "extra-gradient" step which we show stabilizes MD methods by looking ahead and using a "future gradient". These theoretical results are subsequently validated by numerical experiments in GANs.

Point source detection at low signal-to-noise is challenging for astronomical surveys, particularly in radio interferometry images where the noise is correlated. Machine learning is a promising solution, allowing the development of algorithms tailored to specific telescope arrays and science cases. We present DeepSource - a deep learning solution - that uses convolutional neural networks to achieve these goals. DeepSource enhances the Signal-to-Noise Ratio (SNR) of the original map and then uses dynamic blob detection to detect sources. Trained and tested on two sets of 500 simulated 1 deg x 1 deg MeerKAT images with a total of 300,000 sources, DeepSource is essentially perfect in both purity and completeness down to SNR = 4 and outperforms PyBDSF in all metrics. For uniformly-weighted images it achieves a Purity x Completeness (PC) score at SNR = 3 of 0.73, compared to 0.31 for the best PyBDSF model. For natural-weighting we find a smaller improvement of ~40% in the PC score at SNR = 3. If instead we ask where either of the purity or completeness first drop to 90%, we find that DeepSource reaches this value at SNR = 3.6 compared to the 4.3 of PyBDSF (natural-weighting). A key advantage of DeepSource is that it can learn to optimally trade off purity and completeness for any science case under consideration. Our results show that deep learning is a promising approach to point source detection in astronomical images.

Multi-subject fMRI data analysis is an interesting and challenging problem in human brain decoding studies. The inherent anatomical and functional variability across subjects make it necessary to do both anatomical and functional alignment before classification analysis. Besides, when it comes to big data, time complexity becomes a problem that cannot be ignored. This paper proposes Gradient Hyperalignment (Gradient-HA) as a gradient-based functional alignment method that is suitable for multi-subject fMRI datasets with large amounts of samples and voxels. The advantage of Gradient-HA is that it can solve independence and high dimension problems by using Independent Component Analysis (ICA) and Stochastic Gradient Ascent (SGA). Validation using multi-classification tasks on big data demonstrates that Gradient-HA method has less time complexity and better or comparable performance compared with other state-of-the-art functional alignment methods.

The multi-path scattering of light within a pixel [1], bidirectional reflectance distribution [2], and the heterogeneity of sub-pixel constituents [3] are the major concerns in the hyperspectral (HS) data classification. These nonlinearity properties naturally place the HS data on a non-euclidean space. Handling these high dimensional redundant data in a non-euclidean space is one of the major bottlenecks in HS data analysis. Typically, HS classification consists of dimensionality reduction (DR) and subsequent classification operation. The popular DR methods such as principal component analysis (PCA) [4] and linear discriminant analysis (LDA) [5] are linear and operate on Euclidean structures. These linear DR methods skip the curved nonlinear structures of the HS data. On the other hand, manifold learning helps in recovering compact, meaningful low dimensional structures from those complex high dimensional data from a non-euclidean space. The manifold learning methods consider the real world high dimensional data to be generated with a few degrees of freedom [6]. This leads to the projection of the data into lower dimensional space while preserving their underlying geometrical structure [7].

Numerous pattern recognition applications can be formed as learning from graph-structured data, including social network, protein-interaction network, the world wide web data, knowledge graph, etc. While convolutional neural network (CNN) facilitates great advances in gridded image/video understanding tasks, very limited attention has been devoted to transform these successful network structures (including Inception net, Residual net, Dense net, etc.) to establish convolutional networks on graph, due to its irregularity and complexity geometric topologies (unordered vertices, unfixed number of adjacent edges/vertices). In this paper, we aim to give a comprehensive analysis of when work matters by transforming different classical network structures to graph CNN, particularly in the basic graph recognition problem. Specifically, we firstly review the general graph CNN methods, especially in its spectral filtering operation on the irregular graph data. We then introduce the basic structures of ResNet, Inception and DenseNet into graph CNN and construct these network structures on graph, named as G_ResNet, G_Inception, G_DenseNet. In particular, it seeks to help graph CNNs by shedding light on how these classical network structures work and providing guidelines for choosing appropriate graph network frameworks. Finally, we comprehensively evaluate the performance of these different network structures on several public graph datasets (including social networks and bioinformatic datasets), and demonstrate how different network structures work on graph CNN in the graph recognition task.

The new field of biohybrid robotics involves the use of living tissue within robots, rather than just metal and plastic. Muscle is one potential key component of such robots, providing the driving force for movement and function. However, in efforts to integrate living muscle into these machines, there have been problems with the force these muscles can exert and the amount of time before they start to shrink and lose their function. Now, in a study reported in the journal Science Robotics, researchers at The University of Tokyo Institute of Industrial Science have overcome these problems by developing a new method that progresses from individual muscle precursor cells, to muscle-cell-filled sheets, and then to fully functioning skeletal muscle tissues. They incorporated these muscles into a biohybrid robot as antagonistic pairs mimicking those in the body to achieve remarkable robot movement and continued muscle function for over a week.

There are smart speakers, which connect wirelessly to other devices, and then there's the new era of smart speakers, designed to offer services through voice-controlled virtual assistants. Sonos, for a long time, was all about the former, having been a pioneer of high-quality, WiFi-connected speaker systems. Now it's entered the next era with products like Sonos One and Sonos Beam, which work with Amazon's Alexa and other virtual assistants. But Sonos' partners are also rivals, and Sonos' reliance on companies like Amazon, Google, and Apple makes it vulnerable, as it revealed in filing for an initial public offering Friday. To cite one example: Amazon can disable Alexa on Sonos devices any time, with "limited notice," according to the filing.

Daimler will soon take its Mercedes-Benz self-driving cars to the public streets of Beijing. It's the first non-Chinese company to win a license to test level 4 self-driving vehicles there. Level 4 is the second-highest tier of autonomous driving, in which cars can operate without human input in select conditions. Eventually, you might be able to take a nap while these types of vehicles ferry you around. The test vehicles use technology from Daimler's partner Baidu Apollo, and they had to go through rigorous closed-course testing in Beijing and Hebei before Chinese authorities granted the license.

Google's voice-calling "Duplex"- which lets Artificial Intelligence (AI) mimic a human voice to make appointments and book tables through phone calls- may soon enter call centres assisting humans with customer queries. According to a report in The Information late on Thursday, an unnamed insurance company has shown interest in "Duplex" which could "handle simple and repetitive customer calls" before taking help from a human if the conversation gets complicated. Google, however, said in a statement that the company is not testing "Duplex" with any enterprise clients. "We're currently focused on consumer use cases for the'Duplex' technology and we aren't testing'Duplex' with any enterprise clients," a Google spokesperson told Engadget in a statement. "'Duplex is designed to operate in very specific use cases, and currently we're focused on testing with restaurant reservations, hair salon booking and holiday hours with a limited set of trusted testers," the company added. At its annual developer conference in May, Google CEO Sundar Pichai introduced "Duplex" and demonstrated how the AI system could book an appointment at a salon and a table at a restaurant.

Amazon Prime Day is here. The bizarre holiday – which has been running for years and is mostly the concoction of Amazon – offers both great deals and strange offers. Amazon promises the day will be the biggest ever. That is undoubtedly true, but it's important to make sure you don't end up with the biggest buyer's remorse ever, too. It is, really, just a sale.