arXiv subjects

Bin Yang

Publications and source records attributed to Bin Yang.

At least 415 records · Page 23Linked to original sources

arXiv2018

A Machine-learning framework for automatic reference-free quality assessment in MRI

Magnetic resonance (MR) imaging offers a wide variety of imaging techniques. A large amount of data is created per examination which needs to be checked for sufficient quality in order to derive a meaningful diagnosis. This is a manual process and therefore time- and cost-intensive. Any imaging artifacts originating from scanner hardware, signal processing or induced by the patient may reduce the image quality and complicate the diagnosis or any image post-processing. Therefore, the assessment or the ensurance of sufficient image quality in an automated manner is of high interest. Usually no reference image is available or difficult to define. Therefore, classical reference-based approaches are not applicable. Model observers mimicking the human observers (HO) can assist in this task. Thus, we propose a new machine-learning-based reference-free MR image quality assessment framework which is trained on HO-derived labels to assess MR image quality immediately after each acquisition. We include the concept of active learning and present an efficient blinded reading platform to reduce the effort in the HO labeling procedure. Derived image features and the applied classifiers (support-vector-machine, deep neural network) are investigated for a cohort of 250 patients. The MR image quality assessment framework can achieve a high test accuracy of 93.7$\%$ for estimating quality classes on a 5-point Likert-scale. The proposed MR image quality assessment framework is able to provide an accurate and efficient quality estimation which can be used as a prospective quality assurance including automatic acquisition adaptation or guided MR scanner operation, and/or as a retrospective quality assessment including support of diagnostic decisions or quality control in cohort studies.

cs.CV↗

arXiv2018

SBNet: Sparse Blocks Network for Fast Inference

Conventional deep convolutional neural networks (CNNs) apply convolution operators uniformly in space across all feature maps for hundreds of layers - this incurs a high computational cost for real-time applications. For many problems such as object detection and semantic segmentation, we are able to obtain a low-cost computation mask, either from a priori problem knowledge, or from a low-resolution segmentation network. We show that such computation masks can be used to reduce computation in the high-resolution main network. Variants of sparse activation CNNs have previously been explored on small-scale tasks and showed no degradation in terms of object classification accuracy, but often measured gains in terms of theoretical FLOPs without realizing a practical speed-up when compared to highly optimized dense convolution implementations. In this work, we leverage the sparsity structure of computation masks and propose a novel tiling-based sparse convolution algorithm. We verified the effectiveness of our sparse CNN on LiDAR-based 3D object detection, and we report significant wall-clock speed-ups compared to dense convolution without noticeable loss of accuracy.

cs.CV↗

arXiv2018

A New Formulation of The Shortest Path Problem with On-Time Arrival Reliability

We study stochastic routing in the PAth-CEntric (PACE) uncertain road network model. In the PACE model, uncertain travel times are associated with not only edges but also some paths. The uncertain travel times associated with paths are able to well capture the travel time dependency among different edges. This significantly improves the accuracy of travel time distribution estimations for arbitrary paths, which is a fundamental functionality in stochastic routing, compared to classic uncertain road network models where uncertain travel times are associated with only edges. We investigate a new formulation of the shortest path with on-time arrival reliability (SPOTAR) problem under the PACE model. Given a source, a destination, and a travel time budget, the SPOTAR problem aims at finding a path that maximizes the on-time arrival probability. We develop a generic algorithm with different speed-up strategies to solve the SPOTAR problem under the PACE model. Empirical studies with substantial GPS trajectory data offer insight into the design properties of the proposed algorithm and confirm that the algorithm is effective. This report extends the paper "Stochastic Shortest Path Finding in Path-Centric Uncertain Road Networks", to appear in IEEE MDM 2018, by providing a concrete running example of the studied SPOTAR problem in the PACE model and additional statistics of the used GPS trajectories in the experiments.

cs.DB↗

arXiv2018

Multi-level Attention Model for Weakly Supervised Audio Classification

In this paper, we propose a multi-level attention model to solve the weakly labelled audio classification problem. The objective of audio classification is to predict the presence or absence of audio events in an audio clip. Recently, Google published a large scale weakly labelled dataset called Audio Set, where each audio clip contains only the presence or absence of the audio events, without the onset and offset time of the audio events. Our multi-level attention model is an extension to the previously proposed single-level attention model. It consists of several attention modules applied on intermediate neural network layers. The output of these attention modules are concatenated to a vector followed by a multi-label classifier to make the final prediction of each class. Experiments shown that our model achieves a mean average precision (mAP) of 0.360, outperforms the state-of-the-art single-level attention model of 0.327 and Google baseline of 0.314.

eess.AS↗

arXiv2018

Learning to Route with Sparse Trajectory Sets---Extended Version

Motivated by the increasing availability of vehicle trajectory data, we propose learn-to-route, a comprehensive trajectory-based routing solution. Specifically, we first construct a graph-like structure from trajectories as the routing infrastructure. Second, we enable trajectory-based routing given an arbitrary (source, destination) pair. In the first step, given a road network and a collection of trajectories, we propose a trajectory-based clustering method that identifies regions in a road network. If a pair of regions are connected by trajectories, we maintain the paths used by these trajectories and learn a routing preference for travel between the regions. As trajectories are skewed and sparse, many region pairs are not connected by trajectories. We thus transfer routing preferences from region pairs with sufficient trajectories to such region pairs and then use the transferred preferences to identify paths between the regions. In the second step, we exploit the above graph-like structure to achieve a comprehensive trajectory-based routing solution. Empirical studies with two substantial trajectory data sets offer insight into the proposed solution, indicating that it is practical. A comparison with a leading routing service offers evidence that the paper's proposal is able to enhance routing quality. This is an extended version of "Learning to Route with Sparse Trajectory Sets" [1], to appear in IEEE ICDE 2018.

cs.LG↗

arXiv2018

Finding Top-k Optimal Sequenced Routes -- Full Version

Motivated by many practical applications in logistics and mobility-as-a-service, we study the top-k optimal sequenced routes (KOSR) querying on large, general graphs where the edge weights may not satisfy the triangle inequality, e.g., road network graphs with travel times as edge weights. The KOSR querying strives to find the top-k optimal routes (i.e., with the top-k minimal total costs) from a given source to a given destination, which must visit a number of vertices with specific vertex categories (e.g., gas stations, restaurants, and shopping malls) in a particular order (e.g., visiting gas stations before restaurants and then shopping malls). To efficiently find the top-k optimal sequenced routes, we propose two algorithms PruningKOSR and StarKOSR. In PruningKOSR, we define a dominance relationship between two partially-explored routes. The partially-explored routes that can be dominated by other partially-explored routes are postponed being extended, which leads to a smaller searching space and thus improves efficiency. In StarKOSR, we further improve the efficiency by extending routes in an A* manner. With the help of a judiciously designed heuristic estimation that works for general graphs, the cost of partially explored routes to the destination can be estimated such that the qualified complete routes can be found early. In addition, we demonstrate the high extensibility of the proposed algorithms by incorporating Hop Labeling, an effective label indexing technique for shortest path queries, to further improve efficiency. Extensive experiments on multiple real-world graphs demonstrate that the proposed methods significantly outperform the baseline method. Furthermore, when k=1, StarKOSR also outperforms the state-of-the-art method for the optimal sequenced route queries.

cs.DB↗

arXiv2018

Neural Network Ensembles to Real-time Identification of Plug-level Appliance Measurements

The problem of identifying end-use electrical appliances from their individual consumption profiles, known as the appliance identification problem, is a primary stage in both Non-Intrusive Load Monitoring (NILM) and automated plug-wise metering. Therefore, appliance identification has received dedicated studies with various electric appliance signatures, classification models, and evaluation datasets. In this paper, we propose a neural network ensembles approach to address this problem using high resolution measurements. The models are trained on the raw current and voltage waveforms, and thus, eliminating the need for well engineered appliance signatures. We evaluate the proposed model on a publicly available appliance dataset from 55 residential buildings, 11 appliance categories, and over 1000 measurements. We further study the stability of the trained models with respect to training dataset, sampling frequency, and variations in the steady-state operation of appliances.

cs.LG↗

arXiv2018

On the Feasibility of Generic Deep Disaggregation for Single-Load Extraction

Recently, and with the growing development of big energy datasets, data-driven learning techniques began to represent a potential solution to the energy disaggregation problem outperforming engineered and hand-crafted models. However, most proposed deep disaggregation models are load-dependent in the sense that either expert knowledge or a hyper-parameter optimization stage is required prior to training and deployment (normally for each load category) even upon acquisition and cleansing of aggregate and sub-metered data. In this paper, we present a feasibility study on the development of a generic disaggregation model based on data-driven learning. Specifically, we present a generic deep disaggregation model capable of achieving state-of-art performance in load monitoring for a variety of load categories. The developed model is evaluated on the publicly available UK-DALE dataset with a moderately low sampling frequency and various domestic loads.

cs.LG↗

arXiv2018

Selective Sampling and Mixture Models in Generative Adversarial Networks

In this paper, we propose a multi-generator extension to the adversarial training framework, in which the objective of each generator is to represent a unique component of a target mixture distribution. In the training phase, the generators cooperate to represent, as a mixture, the target distribution while maintaining distinct manifolds. As opposed to traditional generative models, inference from a particular generator after training resembles selective sampling from a unique component in the target distribution. We demonstrate the feasibility of the proposed architecture both analytically and with basic Multi-Layer Perceptron (MLP) models trained on the MNIST dataset.

cs.LG↗

arXiv2018

Dense Small Cell Networks: From Noise-Limited to Dense Interference-Limited

Considering both non-line-of-sight (NLoS) and line-of-sight (LoS) transmissions, the transitional behaviors from noise-limited regime to dense interference-limited regime have been investigated for the fifth generation (5G) small cell networks (SCNs). Besides, we identify four performance regimes based on base station (BS) density, i.e., (i) the noise-limited regime, (ii) the signal-dominated regime, (iii) the interference-dominated regime, and (iv) the interference-limited regime. To characterize the performance regime, we propose a unified framework analyzing the future 5G wireless networks over generalized shadowing/fading channels, in which the user association schemes based on the strongest instantaneous received power (SIRP) and the strongest average received power (SARP) can be studied, while NLoS/LoS transmissions and multi-slop path loss model are considered. Simulation results indicate that different factors, i.e., noise, desired signal, and interference, successively and separately dominate the network performance with the increase of BS density. Hence, our results shed new light on the design and management of SCNs in urban and rural areas with different BS deployment densities.

cs.NI↗

arXiv2017

Spectroscopy and thermal modelling of the first interstellar object 1I/2017 U1 'Oumuamua

During the formation and evolution of the Solar System, significant numbers of cometary and asteroidal bodies were ejected into interstellar space$^{1,2}$. It can be reasonably expected that the same happened for planetary systems other than our own. Detection of such Inter- stellar Objects (ISOs) would allow us to probe the planetesimal formation processes around other stars, possibly together with the effects of long-term exposure to the interstellar medium. 1I/2017 U1 'Oumuamua is the first known ISO, discovered by the Pan-STARRS1 telescope in October 2017$^3$.The discovery epoch photometry implies a highly elongated body with radii of $\sim 200 \times 20$ m when a comet-like geometric albedo of 0.04 is assumed. Here we report spectroscopic characterisation of 'Oumuamua, finding it to be variable with time but similar to organically rich surfaces found in the outer Solar System. The observable ISO population is expected to be dominated by comet-like bodies in agreement with our spectra, yet the reported inactivity implies a lack of surface ice. We show this is consistent with predictions of an insulating mantle produced by long-term cosmic ray exposure. An internal icy composition cannot therefore be ruled out by the lack of activity, even though 'Oumuamua passed within 0.25 au of the Sun.

astro-ph.EP↗

arXiv2017

On the Energy-Efficient Deployment for Ultra-Dense Heterogeneous Networks with NLoS and LoS Transmissions

We investigate network performance of ultra-dense heterogeneous networks (HetNets) and study the maximum energy-efficient base station (BS) deployment incorporating probabilistic non-line-of-sight (NLoS) and line-of-sight (LoS) transmissions. First, we develop an analytical framework with the maximum instantaneous received power (MIRP) and the maximum average received power (MARP) association schemes to model the coverage probability and related performance metrics, e.g., the potential throughput (PT) and the energy efficiency (EE). Second, we formulate two optimization problems to achieve the maximum energy-efficient deployment solution with specific service criteria. Simulation results show that there are tradeoffs among the coverage probability, the total power consumption, and the EE. To be specific, the maximum coverage probability with ideal power consumption is superior to that with practical power consumption when the total power constraint is small and inferior to that with practical power consumption when the total power constraint becomes large. Moreover, the maximum EE is a decreasing function with respect to the coverage probability constraint.

cs.NI↗

arXiv2017

Isotopic ratios in outbursting comet C/2015 ER61

Isotopic ratios in comets are critical to understanding the origin of cometary material and the physical and chemical conditions in the early solar nebula. Comet C/2015 ER61 (PANSTARRS) underwent an outburst with a total brightness increase of 2 magnitudes on the night of 2017 April 4. The sharp increase in brightness offered a rare opportunity to measure the isotopic ratios of the light elements in the coma of this comet. We obtained two high-resolution spectra of C/2015 ER61 with UVES/VLT on the nights of 2017 April 13 and 17. At the time of our observations, the comet was fading gradually following the outburst. We measured the nitrogen and carbon isotopic ratios from the CN violet (0,0) band and found that $^{12}$C/$^{13}$C=100 $\pm$ 15, $^{14}$N/$^{15}$N=130 $\pm$ 15. In addition, we determined the $^{14}$N/$^{15}$N ratio from four pairs of NH$_2$ isotopolog lines and measured $^{14}$N/$^{15}$N=140 $\pm$ 28. The measured isotopic ratios of C/2015 ER61 do not deviate significantly from those of other comets.

astro-ph.EP↗

arXiv2017

Reliable Energy-Efficient Routing Algorithm for Vehicle-Assisted Wireless Ad-Hoc Networks

We investigate the design of the optimal routing path in a moving vehicles involved the internet of things (IoT). In our model, jammers exist that may interfere with the information exchange between wireless nodes, leading to worsened quality of service (QoS) in communications. In addition, the transmit power of each battery-equipped node is constrained to save energy. We propose a three-step optimal routing path algorithm for reliable and energy-efficient communications. Moreover, results show that with the assistance of moving vehicles, the total energy consumed can be reduced to a large extend. We also study the impact on the optimal routing path design and energy consumption which is caused by path loss, maximum transmit power constrain, QoS requirement, etc.

cs.NI↗

arXiv2017

The Main Belt Comets and Ice in the Solar System

We review the evidence for buried ice in the asteroid belt; specifically the questions around the so-called Main Belt Comets (MBCs). We summarise the evidence for water throughout the Solar System, and describe the various methods for detecting it, including remote sensing from ultraviolet to radio wavelengths. We review progress in the first decade of study of MBCs, including observations, modelling of ice survival, and discussion on their origins. We then look at which methods will likely be most effective for further progress, including the key challenge of direct detection of (escaping) water in these bodies.

astro-ph.EP↗

arXiv2017

Thermophysical characteristics of the large main-belt asteroid (349) Dembowska

(349) Dembowska, a large, bright main-belt asteroid, has a fast rotation and oblique spin axis. It may have experienced partial melting and differentiation. We constrain Dembowska's thermophysical properties, e.g., thermal inertia, roughness fraction, geometric albedo and effective diameter within 3$σ$ uncertainty of $Γ=20^{+12}_{-7}\rm~Jm^{-2}s^{-0.5}K^{-1}$, $f_{\rm r}=0.25^{+0.60}_{-0.25}$, $p_{\rm v}=0.309^{+0.026}_{-0.038}$, and $D_{\rm eff}=155.8^{+7.5}_{-6.2}\rm~km$, by utilizing the Advanced Thermophysical Model (ATPM) to analyse four sets of thermal infrared data obtained by IRAS, AKARI, WISE and Subaru/COMICS at different epochs. In addition, by modeling the thermal lightcurve observed by WISE, we obtain the rotational phases of each dataset. These rotationally resolved data do not reveal significant variations of thermal inertia and roughness across the surface, indicating the surface of Dembowska should be covered by a dusty regolith layer with few rocks or boulders. Besides, the low thermal inertia of Dembowska show no significant difference with other asteroids larger than 100 km, indicating the dynamical lives of these large asteroids are long enough to make the surface to have sufficiently low thermal inertia. Furthermore, based on the derived surface thermophysical properties, as well as the known orbital and rotational parameters, we can simulate Dembowska's surface and subsurface temperature throughout its orbital period. The surface temperature varies from $\sim40$ K to $\sim220$ K, showing significant seasonal variation, whereas the subsurface temperature achieves equilibrium temperature about $120\sim160$ K below $30\sim50$ cm depth.

astro-ph.EP↗

arXiv2017

T-CNN: Tubelets with Convolutional Neural Networks for Object Detection from Videos

The state-of-the-art performance for object detection has been significantly improved over the past two years. Besides the introduction of powerful deep neural networks such as GoogleNet and VGG, novel object detection frameworks such as R-CNN and its successors, Fast R-CNN and Faster R-CNN, play an essential role in improving the state-of-the-art. Despite their effectiveness on still images, those frameworks are not specifically designed for object detection from videos. Temporal and contextual information of videos are not fully investigated and utilized. In this work, we propose a deep learning framework that incorporates temporal and contextual information from tubelets obtained in videos, which dramatically improves the baseline performance of existing still-image detection frameworks when they are applied to videos. It is called T-CNN, i.e. tubelets with convolutional neueral networks. The proposed framework won the recently introduced object-detection-from-video (VID) task with provided data in the ImageNet Large-Scale Visual Recognition Challenge 2015 (ILSVRC2015).

cs.CV↗

arXiv2017

X-shooter search for outgassing from Main Belt Comet P/2012 T1 (Pan-STARRS)

Main Belt Comets are a recently identified population of minor bodies with stable asteroid-like orbits but cometary appearances. Sublimation of water ice is the most likely mechanism for their recurrent activity (i.e. dust tails and dust comae), although there has been no direct detection of gas. These peculiar objects could hold the key to the origin of water on Earth. In this paper we present a search for the gas responsible for lifting dust from P/2012 T1 (Pan-STARRS), and review previous attempts at such measurements. To date such searches have mainly been indirect, looking for the common cometary gas CN rather than gasses related to water itself. We use the VLT and X-shooter to search for emission from OH in the UV, a direct dissociation product of water. We do not detect any emission lines, and place an upper limit on water production rate from P/2012 T1 of $8-9\times10^{25}$ molecules s$^{-1}$. This is similar to limits derived from observations using the Herschel space telescope. We conclude that the best current facilities are incapable of detecting water emission at the exceptionally low levels required to produce the observed activity in Main Belt Comets.

astro-ph.EP↗