The advances have also fueled speculation of a wave of job losses as machines replace humans, just as the industrial revolution rendered many occupations redundant. The World Economic Forum predicted in January that by 2020, 5 million jobs could be lost to machines. Experts said when it came to the banking and finance sector, the topic needed a more nuanced approach. Speaking on a panel discussing global trends in fintech at InnovFest UnBound, a digital technology conference organized in Singapore, Avinash Hegde, co-founder of a chat bot service Supertext, explained that low-skilled finance jobs, such as basic analytics and number crunching, could soon be done by AI. "The way we interact with business and financial analysts is going to dramatically change in the next few years," he said. Would financial analysts find themselves out of a job?

This tiny robot uses electro-adhesion to "perch" on objects. A Harvard-turned-MIT researcher and his colleagues just dropped some pretty cool Spider-Man tech in the latest issue of Science magazine: surface clinging via "electrostatic adhesion." It's a widely applicable breakthrough that will, for instance, keep future robots perched while they wait for instructions. Flying drones use a lot of energy hovering, but the researchers, who hail from institutes across the U.S. and Hong Kong, may have found the first step in a path to conserving energy during activity. Like birds and insects, drones could save tons of energy if they were able to "perch" instead of hovering, as shown in the video below.

WASHINGTON (Reuters) - Google has no plans to expand its partnership with Fiat Chrysler Automobiles NV to create a self-driving car, the program chief at the Alphabet Inc unit said on Thursday, affirming that the technology company was still in talks with other potential partners. Earlier this month, Google and Fiat Chrysler agreed to work together to build a fleet of 100 self-driving minivans in the most advanced collaboration to date between Silicon Valley and a traditional carmaker. Google said it was not sharing proprietary self-driving vehicle technology with Fiat Chrysler, and that the vehicles would not be offered for sale. "This is just FCA and Google building 100 cars together," Google self-driving car Chief Executive John Krafcik said in an interview on the sidelines of an energy conference in Washington. "We're still talking to a lot of different automakers," he added.

Last month, law firm Baker & Hostetler announced that it would employ IBM's artificially intelligent lawyer, Ross, to help ease its tedious workload. In a statement, the firm's chief technology officer said, "we believe that emerging technologies like cognitive computing and other forms of machine learning can help enhance the services we deliver to our clients." Ross, a system built on the back of IBM's Watson, claims to be able to interpret questions lawyers ask it, and read "through the entire body of law and returns a cited answer and topical readings from legislation, case law and secondary sources to get you up-to-speed quickly." But the first thing I noticed about Ross wasn't how many legal documents it can search at once, or how accurate it claims to be. It was the name: Ross.

The company is working out the bugs in its self-driving technology.Video provided by Newsy Newslook Uber's new self-driving car has begun testing on the streets of Pittsburgh. SAN FRANCISCO -- The first Uber car that doesn't need a driver has hit the streets. The ride-hailing behemoth announced in a blog post Thursday that it has begun testing a self-driving car in Pittsburgh, home of the company's nascent Advanced Technologies Center. The car, a Ford Fusion Hybrid with a roof-full of radar, lasers and cameras, will be collecting road-mapping data as well as testing its real-world traffic reactions. Uber's interest in autonomous car technology dates to a year ago, when the 60 billion start-up began hiring Carnegie Mellon University robotics experts to staff its new center not far from the Pittsburgh-based school.

Wordnets are an effective resource for natural language processing and information retrieval, especially for semantic processing and meaning related tasks. So far, wordnets have been constructed for many languages. However, the automatic development of wordnets for low-resource languages has not been well studied. In this paper, an Expectation-Maximization algorithm is used to create high quality and large scale wordnets for poorresource languages. The proposed method benefits from possessing cross-lingual word sense disambiguation and develops a wordnet by only using a bi-lingual dictionary and a monolingual corpus. The proposed method has been executed with Persian language and the resulting wordnet has been evaluated through several experiments. The results show that the induced wordnet has a precision score of 90% and a recall score of 35%.

In this paper, we present a fully automatic brain tumor segmentation method based on Deep Neural Networks (DNNs). The proposed networks are tailored to glioblastomas (both low and high grade) pictured in MR images. By their very nature, these tumors can appear anywhere in the brain and have almost any kind of shape, size, and contrast. These reasons motivate our exploration of a machine learning solution that exploits a flexible, high capacity DNN while being extremely efficient. Here, we give a description of different model choices that we've found to be necessary for obtaining competitive performance. We explore in particular different architectures based on Convolutional Neural Networks (CNN), i.e. DNNs specifically adapted to image data. We present a novel CNN architecture which differs from those traditionally used in computer vision. Also, different from most traditional uses of CNNs, our networks use a final layer that is a convolutional implementation of a fully connected layer which allows a 40 fold speed up. We also describe a 2-phase training procedure that allows us to tackle difficulties related to the imbalance of tumor labels. Finally, we explore a cascade architecture in which the output of a basic CNN is treated as an additional source of information for a subsequent CNN. Results reported on the 2013 BRATS test dataset reveal that our architecture improves over the currently published state-of-the-art while being over 30 times faster. Keywords: Brain tumor segmentation, deep neural networks 1. Introduction In the United States alone, it is estimated that 23,000 new cases of brain cancer will be diagnosed in 2015 Although surgery is the most common treatment for brain tumors, radiation and chemotherapy may be used to slow the growth of tumors that cannot be physically removed. Magnetic resonance imaging (MRI) provides detailed images of the brain, and is one of the most common tests used to diagnose brain tumors. All the more, brain tumor segmentation from MR images can have great impact for improved diagnostics, growth rate prediction and treatment planning. While some tumors such as meningiomas can be easily segmented, others like gliomas and glioblastomas are much more difficult to localize. Another fundamental difficulty with segmenting brain tumors is that they can appear anywhere in the brain, in almost any shape and size.

Modern computer vision algorithms typically require expensive data acquisition and accurate manual labeling. In this work, we instead leverage the recent progress in computer graphics to generate fully labeled, dynamic, and photo-realistic proxy virtual worlds. We propose an efficient real-to-virtual world cloning method, and validate our approach by building and publicly releasing a new video dataset, called Virtual KITTI (see http://www.xrce.xerox.com/Research-Development/Computer-Vision/Proxy-Virtual-Worlds), automatically labeled with accurate ground truth for object detection, tracking, scene and instance segmentation, depth, and optical flow. We provide quantitative experimental evidence suggesting that (i) modern deep learning algorithms pre-trained on real data behave similarly in real and virtual worlds, and (ii) pre-training on virtual data improves performance. As the gap between real and virtual worlds is small, virtual worlds enable measuring the impact of various weather and imaging conditions on recognition performance, all other things being equal. We show these factors may affect drastically otherwise high-performing deep models for tracking.

We propose an algorithm for detecting patterns exhibited by anomalous clusters in high dimensional discrete data. Unlike most anomaly detection (AD) methods, which detect individual anomalies, our proposed method detects groups (clusters) of anomalies; i.e. sets of points which collectively exhibit abnormal patterns. In many applications this can lead to better understanding of the nature of the atypical behavior and to identifying the sources of the anomalies. Moreover, we consider the case where the atypical patterns exhibit on only a small (salient) subset of the very high dimensional feature space. Individual AD techniques and techniques that detect anomalies using all the features typically fail to detect such anomalies, but our method can detect such instances collectively, discover the shared anomalous patterns exhibited by them, and identify the subsets of salient features. In this paper, we focus on detecting anomalous topics in a batch of text documents, developing our algorithm based on topic models. Results of our experiments show that our method can accurately detect anomalous topics and salient features (words) under each such topic in a synthetic data set and two real-world text corpora and achieves better performance compared to both standard group AD and individual AD techniques. All required code to reproduce our experiments is available from https://github.com/hsoleimani/ATD

We study the problem of sampling k-bandlimited signals on graphs. We propose two sampling strategies that consist in selecting a small subset of nodes at random. The first strategy is non-adaptive, i.e., independent of the graph structure, and its performance depends on a parameter called the graph coherence. On the contrary, the second strategy is adaptive but yields optimal results. Indeed, no more than O(k log(k)) measurements are sufficient to ensure an accurate and stable recovery of all k-bandlimited signals. This second strategy is based on a careful choice of the sampling distribution, which can be estimated quickly. Then, we propose a computationally efficient decoder to reconstruct k-bandlimited signals from their samples. We prove that it yields accurate reconstructions and that it is also stable to noise. Finally, we conduct several experiments to test these techniques.