Søk
Viser treff 1-8 av 8
Ice intelligence retrieval by remote sensing - Possibilities and challenges in an operational setting
(Peer reviewed; Chapter; Bokkapittel, 2017)
Detecting ice drift velocity when operating offshore in ice-covered waters is crucial during marine operations, as ice actions affect station keeping and ice management. Furthermore, other ice data/intelligence such as ice concentration and thickness are important parameters to determine ice resistance, evaluate performance of icebreakers and predict ice actions on structures. Different sensors are ...
Bistatic Observations of the Ocean Surface with HF Radar, Satellite and Airborne Receivers
(Peer reviewed; Book; Bok; Bokkapittel; Chapter, 2017-12-25)
A new concept has been developed which can view vast regions of the Earth's surface. Ground HF transmissions are reflected by the ionosphere to illuminate the ocean over a few thousand kilometers. HF receivers detect the radio waves scattered by the sea and land surface. Using the theory of radio wave scatter from ocean surfaces, the HF data is then processed to yield the directional wave-height ...
Deep kernelized autoencoders
(Peer reviewed; Book; Bokkapittel; Bok; Chapter, 2017-05-19)
In this paper we introduce the deep kernelized autoencoder,
a neural network model that allows an explicit approximation of (i) the
mapping from an input space to an arbitrary, user-specified kernel space
and (ii) the back-projection from such a kernel space to input space. The
proposed method is based on traditional autoencoders and is trained
through a new unsupervised loss function. ...
Assessment of the RISAT-1 FRS-2 Mode for Oil Spill Observation
(Peer reviewed; Chapter, 2017-12-04)
Synthetic aperture radar data acquired by the Radar Imaging Satellite (RISAT-1) over experimental oil spills is here investigated. One quad-polarization scene in the Fine Resolution Alternate Polarization Stripmap (FRS-2) mode is analyzed to evaluate the potential of using this mode for oil spill observation.<br> Oil slicks of varying type and age are clearly detected in the HH and VV channels, with ...
Mission Performance Trade-offs of Battery-powered sUAS
(Chapter; Bokkapittel, 2017-07-27)
A sensitivity analysis is presented on the influence of the weight, altitude and speed of battery-powered sUAS on the resulting stall speed, endurance and range. To aid in the determination of the aircraft performance prior to flight, a method is being brought forth that quantifies the impact of these mission parameters. As a case study the P31015 sUAS is used. The P31015 is a concept model of a ...
Temporal overdrive recurrent neural network
(Chapter; Bokkapittel, 2017-07-03)
In this work we present a novel recurrent neural network architecture designed to model systems characterized by multiple characteristic timescales in their dynamics. The proposed network is composed by several recurrent groups of neurons that are trained to separately adapt to each timescale, in order to improve the system identification process. We test our framework on time series prediction tasks ...
Critical echo state network dynamics by means of Fisher information maximization
(Chapter; Bokkapittel, 2017-07-03)
The computational capability of an Echo State Network (ESN), expressed in terms of low prediction error and high short-term memory capacity, is maximized on the so-called “edge of criticality”. In this paper we present a novel, unsupervised approach to identify this edge and, accordingly, we determine hyperparameters configuration that maximize network performance. The proposed method is ...
Large-Scale Mapping of Small Roads in Lidar Images Using Deep Convolutional Neural Networks
(Chapter; Bokkapittel, 2017-05-19)
Detailed and complete mapping of forest roads is important for the forest industry since they are used for timber transport by trucks with long trailers. This paper proposes a new automatic method for large-scale mapping forest roads from airborne laser scanning data. The method is based on a fully convolutional neural network that performs end-to-end segmentation. To train the network, a large set ...