Collaborating Authors


'Largest drone war in the world': How airpower saved Tripoli

Al Jazeera

Air power has played an increasingly important role in the Libyan conflict. The relatively flat featureless desert terrain of the north and coast means that ground units are easily spotted, with few places to hide. The air forces of both the United Nations-recognised Government of National Accord (GNA) and eastern-based commander Khalifa Haftar and his self-styled Libyan National Army (LNA) use French and Soviet-era fighter jets, antiquated and poorly maintained. While manned fighter aircraft have been used, for the most part the air war has been fought by unmanned aerial vehicles (UAVs) or drones. With nearly 1,000 air strikes conducted by UAVs, UN Special Representative to Libya Ghassan Salame called the conflict "the largest drone war in the world".

Alternating Between Spectral and Spatial Estimation for Speech Separation and Enhancement Machine Learning

This work investigates alternation between spectral separation using masking-based networks and spatial separation using multichannel beamforming. In this framework, the spectral separation is performed using a mask-based deep network. The result of mask-based separation is used, in turn, to estimate a spatial beamformer. The output of the beamformer is fed back into another mask-based separation network. We explore multiple ways of computing time-varying covariance matrices to improve beamforming, including factorizing the spatial covariance into a time-varying amplitude component and time-invariant spatial component. For the subsequent mask-based filtering, we consider different modes, including masking the noisy input, masking the beamformer output, and a hybrid approach combining both. Our best method first uses spectral separation, then spatial beamforming, and finally a spectral post-filter, and demonstrates an average improvement of 2.8 dB over baseline mask-based separation, across four different reverberant speech enhancement and separation tasks.

How Sensors, Robotics And Artificial Intelligence Will Transform Agriculture


"Plant breeding is another interesting application we're pursuing, where robotically gathered plant phenotype data can be collected over much larger breeding experiments that current manual measurement techniques allow," said Kantor. "Machine learning tools can then combine the collected phenotype data with genetic and environmental data to help a breeders and geneticists better understand the relationships between genetics, environment, and plant performance." "This in turn accelerates the breeding process, allowing breeders to evaluate many more plants each season so that they can more quickly select for desirable traits such as yield or disease resistance," adds Kantor. Kantor says this kind of accelerated breeding program could have significant benefit in the developing world such as Subsaharan Africa. The FarmView initiative wants to develop inexpensive robotic systems that small- to medium-scale growers can afford to implement.

Flynn: Outspoken general, intelligence pro, Trump supporter

Associated Press

Retired Lt. Gen Michael Flynn gestures as he arrives at Trump Tower, Thursday, Nov. 17, 2016, in New York. Retired Lt. Gen Michael Flynn gestures as he arrives at Trump Tower, Thursday, Nov. 17, 2016, in New York. FILE- In this file photo taken on Thursday, Dec. 10, 2015, Russian President Vladimir Putin, center right, with retired U.S. Lt. Gen. Michael T. Flynn, center left, and Serbian filmmaker Emir Kusturica, obscured second right, attend an exhibition marking the 10th anniversary of RT (Russia Today) 24-hour English-language TV news channel in Moscow, Russia. WASHINGTON (AP) -- Michael Flynn, the former Army lieutenant general Donald Trump has asked to be his national security adviser, rose through the ranks of military intelligence on the strength of his reputation as an astute professional and an unconventional thinker.


AAAI Conferences

The RoboCup robot soccer Small Size League has been running since 1997 with many teams successfully competiting and very effectively playing the games. Teams of five robots, with a combined autonomous centralized perception and control, and distributed actuation, move at high speeds in the field space, actuating a golf ball by passing and shooting it to aim at scoring goals. Most teams run their own pre-defined team strategies, unknown to the other teams, with flexible game-state dependent assignment of robot roles and positioning. However, in this fast-paced noisy real robot league, recognizing the opponent team strategies and accordingly adapting one's own play has proven to be a considerable challenge. In this work, we analyze logged data of real games gathered by the CMDragons team, and contribute several results in learning and responding to opponent strategies. We define episodes as segments of interest in the logged data, and introduce a representation that captures the spatial and temporal data of the multi-robot system as instances of geometrical trajectory curves. We then learn a model of the team strategies through a variant of agglomerative hierarchical clustering. Using the learned cluster model, we are able to classify a team behavior incrementally as it occurs. Finally, we define an algorithm that autonomously generates counter tactics, in a simulation based on the real logs, showing that it can recognize and respond to opponent strategies.


AAAI Conferences

We introduce a new method to find semantic inconsistencies (i.e., concepts with erroneous synonymity) in the Unified Medical Language System (UMLS). The idea is to identify the inconsistencies by comparing the semantic groups of hierarchically-related concepts using Answer Set Programming. With this method, we identified several inconsistent concepts in UMLS and discovered an interesting semantic pattern along hierarchies, which seems associated with wrong synonymy.