##### Browse all papers

Proceedings ArticleDOI
20 Mar 2017-
TL;DR: This work presents a conceptually simple, flexible, and general framework for object instance segmentation, which extends Faster R-CNN by adding a branch for predicting an object mask in parallel with the existing branch for bounding box recognition.

9,492 citations

Proceedings ArticleDOI
, Jian Sun1
07 Dec 2015-
Abstract: Rectified activation units (rectifiers) are essential for state-of-the-art neural networks. In this work, we study rectifier neural networks for image classification from two aspects. First, we propose a Parametric Rectified Linear Unit (PReLU) that generalizes the traditional rectified unit. PReLU improves model fitting with nearly zero extra computational cost and little overfitting risk. Second, we derive a robust initialization method that particularly considers the rectifier nonlinearities. This method enables us to train extremely deep rectified models directly from scratch and to investigate deeper or wider network architectures. Based on the learnable activation and advanced initialization, we achieve 4.94% top-5 test error on the ImageNet 2012 classification dataset. This is a 26% relative improvement over the ILSVRC 2014 winner (GoogLeNet, 6.66% [33]). To our knowledge, our result is the first to surpass the reported human-level performance (5.1%, [26]) on this dataset.

9,436 citations

Journal ArticleDOI
Abstract: We present results based on full-mission Planck observations of temperature and polarization anisotropies of the CMB. These data are consistent with the six-parameter inflationary LCDM cosmology. From the Planck temperature and lensing data, for this cosmology we find a Hubble constant, H0= (67.8 +/- 0.9) km/s/Mpc, a matter density parameter Omega_m = 0.308 +/- 0.012 and a scalar spectral index with n_s = 0.968 +/- 0.006. (We quote 68% errors on measured parameters and 95% limits on other parameters.) Combined with Planck temperature and lensing data, Planck LFI polarization measurements lead to a reionization optical depth of tau = 0.066 +/- 0.016. Combining Planck with other astrophysical data we find N_ eff = 3.15 +/- 0.23 for the effective number of relativistic degrees of freedom and the sum of neutrino masses is constrained to < 0.23 eV. Spatial curvature is found to be |Omega_K| < 0.005. For LCDM we find a limit on the tensor-to-scalar ratio of r <0.11 consistent with the B-mode constraints from an analysis of BICEP2, Keck Array, and Planck (BKP) data. Adding the BKP data leads to a tighter constraint of r < 0.09. We find no evidence for isocurvature perturbations or cosmic defects. The equation of state of dark energy is constrained to w = -1.006 +/- 0.045. Standard big bang nucleosynthesis predictions for the Planck LCDM cosmology are in excellent agreement with observations. We investigate annihilating dark matter and deviations from standard recombination, finding no evidence for new physics. The Planck results for base LCDM are in agreement with BAO data and with the JLA SNe sample. However the amplitude of the fluctuations is found to be higher than inferred from rich cluster counts and weak gravitational lensing. Apart from these tensions, the base LCDM cosmology provides an excellent description of the Planck CMB observations and many other astrophysical data sets.

9,349 citations

Posted Content
TL;DR: The TensorFlow interface and an implementation of that interface that is built at Google are described, which has been used for conducting research and for deploying machine learning systems into production across more than a dozen areas of computer science and other fields.
Abstract: TensorFlow is an interface for expressing machine learning algorithms, and an implementation for executing such algorithms. A computation expressed using TensorFlow can be executed with little or no change on a wide variety of heterogeneous systems, ranging from mobile devices such as phones and tablets up to large-scale distributed systems of hundreds of machines and thousands of computational devices such as GPU cards. The system is flexible and can be used to express a wide variety of algorithms, including training and inference algorithms for deep neural network models, and it has been used for conducting research and for deploying machine learning systems into production across more than a dozen areas of computer science and other fields, including speech recognition, computer vision, robotics, information retrieval, natural language processing, geographic information extraction, and computational drug discovery. This paper describes the TensorFlow interface and an implementation of that interface that we have built at Google. The TensorFlow API and a reference implementation were released as an open-source package under the Apache 2.0 license in November, 2015 and are available at www.tensorflow.org.

9,253 citations

Proceedings ArticleDOI
21 Jul 2017-
TL;DR: Conditional adversarial networks are investigated as a general-purpose solution to image-to-image translation problems and it is demonstrated that this approach is effective at synthesizing photos from label maps, reconstructing objects from edge maps, and colorizing images, among other tasks.
Abstract: We investigate conditional adversarial networks as a general-purpose solution to image-to-image translation problems. These networks not only learn the mapping from input image to output image, but also learn a loss function to train this mapping. This makes it possible to apply the same generic approach to problems that traditionally would require very different loss formulations. We demonstrate that this approach is effective at synthesizing photos from label maps, reconstructing objects from edge maps, and colorizing images, among other tasks. Moreover, since the release of the pix2pix software associated with this paper, hundreds of twitter users have posted their own artistic experiments using our system. As a community, we no longer hand-engineer our mapping functions, and this work suggests we can achieve reasonable results without handengineering our loss functions either.

9,134 citations

Posted Content
, Jian Sun1
TL;DR: This work proposes a Parametric Rectified Linear Unit (PReLU) that generalizes the traditional rectified unit and derives a robust initialization method that particularly considers the rectifier nonlinearities.
Abstract: Rectified activation units (rectifiers) are essential for state-of-the-art neural networks. In this work, we study rectifier neural networks for image classification from two aspects. First, we propose a Parametric Rectified Linear Unit (PReLU) that generalizes the traditional rectified unit. PReLU improves model fitting with nearly zero extra computational cost and little overfitting risk. Second, we derive a robust initialization method that particularly considers the rectifier nonlinearities. This method enables us to train extremely deep rectified models directly from scratch and to investigate deeper or wider network architectures. Based on our PReLU networks (PReLU-nets), we achieve 4.94% top-5 test error on the ImageNet 2012 classification dataset. This is a 26% relative improvement over the ILSVRC 2014 winner (GoogLeNet, 6.66%). To our knowledge, our result is the first to surpass human-level performance (5.1%, Russakovsky et al.) on this visual recognition challenge.

8,865 citations

Journal ArticleDOI
16 Sep 2017-The Lancet
Abstract: Summary Background As mortality rates decline, life expectancy increases, and populations age, non-fatal outcomes of diseases and injuries are becoming a larger component of the global burden of disease. The Global Burden of Diseases, Injuries, and Risk Factors Study 2016 (GBD 2016) provides a comprehensive assessment of prevalence, incidence, and years lived with disability (YLDs) for 328 causes in 195 countries and territories from 1990 to 2016. Methods We estimated prevalence and incidence for 328 diseases and injuries and 2982 sequelae, their non-fatal consequences. We used DisMod-MR 2.1, a Bayesian meta-regression tool, as the main method of estimation, ensuring consistency between incidence, prevalence, remission, and cause of death rates for each condition. For some causes, we used alternative modelling strategies if incidence or prevalence needed to be derived from other data. YLDs were estimated as the product of prevalence and a disability weight for all mutually exclusive sequelae, corrected for comorbidity and aggregated to cause level. We updated the Socio-demographic Index (SDI), a summary indicator of income per capita, years of schooling, and total fertility rate. GBD 2016 complies with the Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER). Findings Globally, low back pain, migraine, age-related and other hearing loss, iron-deficiency anaemia, and major depressive disorder were the five leading causes of YLDs in 2016, contributing 57·6 million (95% uncertainty interval [UI] 40·8–75·9 million [7·2%, 6·0–8·3]), 45·1 million (29·0–62·8 million [5·6%, 4·0–7·2]), 36·3 million (25·3–50·9 million [4·5%, 3·8–5·3]), 34·7 million (23·0–49·6 million [4·3%, 3·5–5·2]), and 34·1 million (23·5–46·0 million [4·2%, 3·2–5·3]) of total YLDs, respectively. Age-standardised rates of YLDs for all causes combined decreased between 1990 and 2016 by 2·7% (95% UI 2·3–3·1). Despite mostly stagnant age-standardised rates, the absolute number of YLDs from non-communicable diseases has been growing rapidly across all SDI quintiles, partly because of population growth, but also the ageing of populations. The largest absolute increases in total numbers of YLDs globally were between the ages of 40 and 69 years. Age-standardised YLD rates for all conditions combined were 10·4% (95% UI 9·0–11·8) higher in women than in men. Iron-deficiency anaemia, migraine, Alzheimer's disease and other dementias, major depressive disorder, anxiety, and all musculoskeletal disorders apart from gout were the main conditions contributing to higher YLD rates in women. Men had higher age-standardised rates of substance use disorders, diabetes, cardiovascular diseases, cancers, and all injuries apart from sexual violence. Globally, we noted much less geographical variation in disability than has been documented for premature mortality. In 2016, there was a less than two times difference in age-standardised YLD rates for all causes between the location with the lowest rate (China, 9201 YLDs per 100 000, 95% UI 6862–11943) and highest rate (Yemen, 14 774 YLDs per 100 000, 11 018–19 228). Interpretation The decrease in death rates since 1990 for most causes has not been matched by a similar decline in age-standardised YLD rates. For many large causes, YLD rates have either been stagnant or have increased for some causes, such as diabetes. As populations are ageing, and the prevalence of disabling disease generally increases steeply with age, health systems will face increasing demand for services that are generally costlier than the interventions that have led to declines in mortality in childhood or for the major causes of mortality in adults. Up-to-date information about the trends of disease and how this varies between countries is essential to plan for an adequate health-system response. Funding Bill & Melinda Gates Foundation, and the National Institute on Aging and the National Institute of Mental Health of the National Institutes of Health.

8,768 citations

Journal ArticleDOI

26 Jan 2017-Scientific Reports
TL;DR: It is found that intraflagellar transport 20 mediates the ability of Ror2 signaling to induce the invasiveness of tumors that lack primary cilia, and IFT20 regulates the nucleation of Golgi-derived microtubules by affecting the GM130-AKAP450 complex.
Abstract: Signaling through the Ror2 receptor tyrosine kinase promotes invadopodia formation for tumor invasion. Here, we identify intraflagellar transport 20 (IFT20) as a new target of this signaling in tumors that lack primary cilia, and find that IFT20 mediates the ability of Ror2 signaling to induce the invasiveness of these tumors. We also find that IFT20 regulates the nucleation of Golgi-derived microtubules by affecting the GM130-AKAP450 complex, which promotes Golgi ribbon formation in achieving polarized secretion for cell migration and invasion. Furthermore, IFT20 promotes the efficiency of transport through the Golgi complex. These findings shed new insights into how Ror2 signaling promotes tumor invasiveness, and also advance the understanding of how Golgi structure and transport can be regulated.

8,752 citations

Journal ArticleDOI
TL;DR: This document provides updated normal values for all four cardiac chambers, including three-dimensional echocardiography and myocardial deformation, when possible, on the basis of considerably larger numbers of normal subjects, compiled from multiple databases.
Abstract: The rapid technological developments of the past decade and the changes in echocardiographic practice brought about by these developments have resulted in the need for updated recommendations to the previously published guidelines for cardiac chamber quantification, which was the goal of the joint writing group assembled by the American Society of Echocardiography and the European Association of Cardiovascular Imaging. This document provides updated normal values for all four cardiac chambers, including three-dimensional echocardiography and myocardial deformation, when possible, on the basis of considerably larger numbers of normal subjects, compiled from multiple databases. In addition, this document attempts to eliminate several minor discrepancies that existed between previously published guidelines.

8,690 citations

Proceedings ArticleDOI
01 Oct 2017-
Abstract: Image-to-image translation is a class of vision and graphics problems where the goal is to learn the mapping between an input image and an output image using a training set of aligned image pairs. However, for many tasks, paired training data will not be available. We present an approach for learning to translate an image from a source domain X to a target domain Y in the absence of paired examples. Our goal is to learn a mapping G : X → Y such that the distribution of images from G(X) is indistinguishable from the distribution Y using an adversarial loss. Because this mapping is highly under-constrained, we couple it with an inverse mapping F : Y → X and introduce a cycle consistency loss to push F(G(X)) ≈ X (and vice versa). Qualitative results are presented on several tasks where paired training data does not exist, including collection style transfer, object transfiguration, season transfer, photo enhancement, etc. Quantitative comparisons against several prior methods demonstrate the superiority of our approach.

8,605 citations

Journal ArticleDOI
TL;DR: Slow momentum for some cancers amenable to early detection is juxtaposed with notable gains for other common cancers, and it is notable that long‐term rapid increases in liver cancer mortality have attenuated in women and stabilized in men.
Abstract: Each year, the American Cancer Society estimates the numbers of new cancer cases and deaths that will occur in the United States and compiles the most recent data on population-based cancer occurrence. Incidence data (through 2016) were collected by the Surveillance, Epidemiology, and End Results Program; the National Program of Cancer Registries; and the North American Association of Central Cancer Registries. Mortality data (through 2017) were collected by the National Center for Health Statistics. In 2020, 1,806,590 new cancer cases and 606,520 cancer deaths are projected to occur in the United States. The cancer death rate rose until 1991, then fell continuously through 2017, resulting in an overall decline of 29% that translates into an estimated 2.9 million fewer cancer deaths than would have occurred if peak rates had persisted. This progress is driven by long-term declines in death rates for the 4 leading cancers (lung, colorectal, breast, prostate); however, over the past decade (2008-2017), reductions slowed for female breast and colorectal cancers, and halted for prostate cancer. In contrast, declines accelerated for lung cancer, from 3% annually during 2008 through 2013 to 5% during 2013 through 2017 in men and from 2% to almost 4% in women, spurring the largest ever single-year drop in overall cancer mortality of 2.2% from 2016 to 2017. Yet lung cancer still caused more deaths in 2017 than breast, prostate, colorectal, and brain cancers combined. Recent mortality declines were also dramatic for melanoma of the skin in the wake of US Food and Drug Administration approval of new therapies for metastatic disease, escalating to 7% annually during 2013 through 2017 from 1% during 2006 through 2010 in men and women aged 50 to 64 years and from 2% to 3% in those aged 20 to 49 years; annual declines of 5% to 6% in individuals aged 65 years and older are particularly striking because rates in this age group were increasing prior to 2013. It is also notable that long-term rapid increases in liver cancer mortality have attenuated in women and stabilized in men. In summary, slowing momentum for some cancers amenable to early detection is juxtaposed with notable gains for other common cancers.

8,578 citations

Posted Content
TL;DR: This work shows that it can significantly improve the acoustic model of a heavily used commercial system by distilling the knowledge in an ensemble of models into a single model and introduces a new type of ensemble composed of one or more full models and many specialist models which learn to distinguish fine-grained classes that the full models confuse.
Abstract: A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions. Unfortunately, making predictions using a whole ensemble of models is cumbersome and may be too computationally expensive to allow deployment to a large number of users, especially if the individual models are large neural nets. Caruana and his collaborators have shown that it is possible to compress the knowledge in an ensemble into a single model which is much easier to deploy and we develop this approach further using a different compression technique. We achieve some surprising results on MNIST and we show that we can significantly improve the acoustic model of a heavily used commercial system by distilling the knowledge in an ensemble of models into a single model. We also introduce a new type of ensemble composed of one or more full models and many specialist models which learn to distinguish fine-grained classes that the full models confuse. Unlike a mixture of experts, these specialist models can be trained rapidly and in parallel.

8,473 citations

Journal ArticleDOI
TL;DR: Quantitative assessments show that SegNet provides good performance with competitive inference time and most efficient inference memory-wise as compared to other architectures, including FCN and DeconvNet.
Abstract: We present a novel and practical deep fully convolutional neural network architecture for semantic pixel-wise segmentation termed SegNet. This core trainable segmentation engine consists of an encoder network, a corresponding decoder network followed by a pixel-wise classification layer. The architecture of the encoder network is topologically identical to the 13 convolutional layers in the VGG16 network [1] . The role of the decoder network is to map the low resolution encoder feature maps to full input resolution feature maps for pixel-wise classification. The novelty of SegNet lies is in the manner in which the decoder upsamples its lower resolution input feature map(s). Specifically, the decoder uses pooling indices computed in the max-pooling step of the corresponding encoder to perform non-linear upsampling. This eliminates the need for learning to upsample. The upsampled maps are sparse and are then convolved with trainable filters to produce dense feature maps. We compare our proposed architecture with the widely adopted FCN [2] and also with the well known DeepLab-LargeFOV [3] , DeconvNet [4] architectures. This comparison reveals the memory versus accuracy trade-off involved in achieving good segmentation performance. SegNet was primarily motivated by scene understanding applications. Hence, it is designed to be efficient both in terms of memory and computational time during inference. It is also significantly smaller in the number of trainable parameters than other competing architectures and can be trained end-to-end using stochastic gradient descent. We also performed a controlled benchmark of SegNet and other architectures on both road scenes and SUN RGB-D indoor scene segmentation tasks. These quantitative assessments show that SegNet provides good performance with competitive inference time and most efficient inference memory-wise as compared to other architectures. We also provide a Caffe implementation of SegNet and a web demo at http://mi.eng.cam.ac.uk/projects/segnet/ .

8,450 citations

Book ChapterDOI
Abstract: The concept of one event happening before another in a distributed system is examined, and is shown to define a partial ordering of the events. A distributed algorithm is given for synchronizing a system of logical clocks which can be used to totally order the events. The use of the total ordering is illustrated with a method for solving synchronization problems. The algorithm is then specialized for synchronizing physical clocks, and a bound is derived on how far out of synchrony the clocks can become.

8,352 citations

Proceedings ArticleDOI
07 Jun 2015-
TL;DR: A system that directly learns a mapping from face images to a compact Euclidean space where distances directly correspond to a measure offace similarity, and achieves state-of-the-art face recognition performance using only 128-bytes perface.
Abstract: Despite significant recent advances in the field of face recognition [10, 14, 15, 17], implementing face verification and recognition efficiently at scale presents serious challenges to current approaches. In this paper we present a system, called FaceNet, that directly learns a mapping from face images to a compact Euclidean space where distances directly correspond to a measure of face similarity. Once this space has been produced, tasks such as face recognition, verification and clustering can be easily implemented using standard techniques with FaceNet embeddings as feature vectors.

8,289 citations

Posted Content
09 Sep 2016-arXiv: Learning
TL;DR: A scalable approach for semi-supervised learning on graph-structured data that is based on an efficient variant of convolutional neural networks which operate directly on graphs which outperforms related methods by a significant margin.
Abstract: We present a scalable approach for semi-supervised learning on graph-structured data that is based on an efficient variant of convolutional neural networks which operate directly on graphs. We motivate the choice of our convolutional architecture via a localized first-order approximation of spectral graph convolutions. Our model scales linearly in the number of graph edges and learns hidden layer representations that encode both local graph structure and features of nodes. In a number of experiments on citation networks and on a knowledge graph dataset we demonstrate that our approach outperforms related methods by a significant margin.

8,285 citations

Journal ArticleDOI
TL;DR: The 2016 World Health Organization Classification of Tumors of the Central Nervous System is both a conceptual and practical advance over its 2007 predecessor and is hoped that it will facilitate clinical, experimental and epidemiological studies that will lead to improvements in the lives of patients with brain tumors.
Abstract: The 2016 World Health Organization Classification of Tumors of the Central Nervous System is both a conceptual and practical advance over its 2007 predecessor. For the first time, the WHO classification of CNS tumors uses molecular parameters in addition to histology to define many tumor entities, thus formulating a concept for how CNS tumor diagnoses should be structured in the molecular era. As such, the 2016 CNS WHO presents major restructuring of the diffuse gliomas, medulloblastomas and other embryonal tumors, and incorporates new entities that are defined by both histology and molecular features, including glioblastoma, IDH-wildtype and glioblastoma, IDH-mutant; diffuse midline glioma, H3 K27M-mutant; RELA fusion-positive ependymoma; medulloblastoma, WNT-activated and medulloblastoma, SHH-activated; and embryonal tumour with multilayered rosettes, C19MC-altered. The 2016 edition has added newly recognized neoplasms, and has deleted some entities, variants and patterns that no longer have diagnostic and/or biological relevance. Other notable changes include the addition of brain invasion as a criterion for atypical meningioma and the introduction of a soft tissue-type grading system for the now combined entity of solitary fibrous tumor / hemangiopericytoma-a departure from the manner by which other CNS tumors are graded. Overall, it is hoped that the 2016 CNS WHO will facilitate clinical, experimental and epidemiological studies that will lead to improvements in the lives of patients with brain tumors.

8,183 citations

Journal ArticleDOI

01 Sep 2015-SoftwareX
TL;DR: GROMACS is one of the most widely used open-source and free software codes in chemistry, used primarily for dynamical simulations of biomolecules, and provides a rich set of calculation types.
Abstract: GROMACS is one of the most widely used open-source and free software codes in chemistry, used primarily for dynamical simulations of biomolecules. It provides a rich set of calculation types, prepa ...

8,050 citations

Journal ArticleDOI
, A. Ain10, P. Ajith11, P. A. Altin13, J. S. Areeda16, N. Arnaud17, K. G. Arun18, S. Ascenzi19, M. Ast21, P. Bacon22, P. T. Baker23, B. Barr25, M. Barsuglia22, Imre Bartos26, A. Basti3, C. Baune7, A. S. Bell25, Chris J Bell25, R. Birney31, Carl Blair32, David Blair32, O. Bock7, M. Boer5, C. Bogan7, P. Bojtos34, R. Bork1, J. E. Brau36, V. Brisson17, P. Brockill14, D. D. Brown29, Tomasz Bulik28, H. J. Bulten37, C. Buy22, Jordan Camp43, Junwei Cao44, E. Capocasa22, S. Caride45, C. Casentini19, F. Cavalier17, E. Cesarini19, M. Chan25, Shiuh Chao47, P. Charlton48, Yi Chen1, H. S. Cho50, M. Cho38, Jong H. Chow13, Q. Chu32, S. Chung32, F. Cleva5, E. Coccia19, A. Colla52, A. Conte52, L. Conti3, E. E. Cowan39, David Coward32, R. Coyne45, Kieran Craig25, A. M. Cruise29, A. Cumming25, N. S. Darman58, M. Davier17, Gavin Davies25, E. J. Daw59, Soumi De24, D.B. DeBra27, W. Del Pozzo29, R. De Rosa42, S. Di Pace52, V. Dolique40, K. L. Dooley15, R. Douglas25, T. P. Downes14, M. Drago3, Zhihui Du44, T. B. Edo59, T. Etzel1, R. Everett46, V. Fafone19, H. Fair24, Xiaohui Fan44, Qi Fang32, Ben Farr49, Will M. Farr29, Marc Favata60, M. Fays61, D. Fiorucci22, M. Fletcher25, S. Franco17, S. Frasca52, Z. Frei34, R. Frey36, V. Frey17, F. Garufi42, Neil Gehrels43, Sourav Ghosh33, Karl Gill64, A. Glaefke25, Evan Goetz65, R. Goetz4, N. A. Gordon25, C. Graef25, P. B. Graff38, M. Granata40, A. Grant25, G. Greco3, X. Guo44, A. Gupta10, R. Gustafson65, J. J. Hacker16, B. R. Hall35, M. Haney11, Chad Hanna46, Mark Hannam61, M. J. Hart25, K. Haughian25, James Healy68, J. Hennig25, Stefan Hild25, D. Hoak69, D. J. Hosken70, J. H. Hough25, Eric Howell32, Yi-Ming Hu25, S. Huang47, E. A. Huerta71, D. Huet17, B. Hughey64, Sascha Husa41, A. Idrisy46, N. Indik7, R. Inta45, H. N. Isa25, G. Islas16, Bala R. Iyer11, H. J. Jang50, Karan Jani39, S. Jawahar73, David Jones20, Roger Jones25, Li Ju32, K. Haris74, V. Kalogera55, G. Kang50, S. Karki36, David Keitel41, D. B. Kelley24, W. Kells1, R. Kennedy59, J. S. Key56, Jinsook Kim76, Kyungmin Kim77, Nam Gyu Kim50, Namjun Kim27, Y. M. Kim76, E. J. King70, L. Kleybolte21, M. Korobko21, I. Kowalska28, G. Kuehn7, Rajesh Kumar25, L. Kuo47, P. Kwee7, B. D. Lackey24, J. S. Lange68, B. Lantz27, C. Lazzaro39, P. Leaci52, S. Leavey25, Hyun Lee77, Kejia Lee25, A. Lenon24, N. Leroy17, Yuri Levin78, J. Logue25, J. E. Lord24, J. Luo51, Y. Q. Ma32, R. M. Magee35, V. Malvezzi19, N. Man5, Ilya Mandel29, Vuk Mandic57, V. Mangano25, Zsuzsa Márka26, E. Maros1, I. W. Martin25, K. Mason8, L. Matone26, N. Mazumder35, D. Meacher46, R. A. Mercer14, C. Messenger25, Cody Messick46, P. M. Meyers57, F. Mezzani52, Haixing Miao29, C. Michel40, H. Middleton29, J. Ming7, A. Moggi3, Blake Moore60, B. Mours5, A. W. Muir61, N. Mukund10, Jesper Munch70, David Murphy26, P. G. Murray25, R. K. Nayak85, K. Nedkova69, M. Neri3, A. Neunzert65, G. P. Newton25, T. T. Nguyen13, Evan Ochsner14, J. O'Dell66, G. H. Ogin86, F. Ohme61, M. Oliver41, Archana Pai74, A. Pal-Singh21, Howard Pan47, Yi Pan38, Chris Pankow55, H. R. Paris27, D. Pascucci25, Z. Patrick27, R. Pedurand40, S. Penn87, M. Phelps25, L. Pinard40, J. Poeld7, A. Post7, Jade Powell25, J. Prasad10, V. Predoi61, L. Prokhorov30, P. Puppo3, H. Qi14, Jiayi Qin32, V. Quetschke56, P. Raffai34, P. Rapagnani52, V. Re19, L. Rei3, Stuart Reid31, H. Rew82, Susana Reyes24, F. Ricci52, K. Riles65, R. Robie25, F. Robinet17, V. J. Roma36, Gleb Romanov82, D. Rosińska88, Sheila Rowan25, M. Saleem74, A. Samajdar85, L. Sammut78, B. Sandeen55, B. Sassolas40, P. Schale36, A. Schönbeck21, V. Sequino19, G. Serna16, Y. Setyawati33, Sweta Shah33, Z. Shao1, B. Shapiro27, P. Shawhan38, A. Sheperd14, K. Siellez39, Leo Singer43, J. R. Smith16, B. Sorazu25, A. Staley26, Robert Stone56, N. Straniero40, R. Sturani83, L. Sun58, P. J. Sutton61, M. Tacca22, D. Talukder36, Márton Tápai63, T. Theeg7, Eric Thrane78, V. Tiwari61, C. Tomlinson59, C. V. Torres56, D. Töyrä29, L. Trozzo91, D. Ugolini92, A. L. Urban14, S. A. Usman24, G. Valdes56, S. Vass1, John Veitch29, P. J. Veitch70, D. J. Vine31, T. Vo24, H. Vocca3, A. R. Wade13, L. E. Wade94, S. Walsh7, G. Wang3, Hua Wang29, Meng Wang29, X. F. Wang44, Yanzhi Wang32, H. Ward25, R. L. Ward13, M. Was5, L.-W. Wei95, Linqing Wen32, K. Wette7, James Whelan68, D. J. White59, Graham Woan25, J. L. Wright25, C. C. Yancey38, M. J. Yap13, M. Yvert5, M. Zevin55, Mi Zhang82, Y.-H. Zhang68, C. Zhao32, Zifan Zhou55, S. E. Zuraw69
TL;DR: This is the first direct detection of gravitational waves and the first observation of a binary black hole merger, and these observations demonstrate the existence of binary stellar-mass black hole systems.
Abstract: On September 14, 2015 at 09:50:45 UTC the two detectors of the Laser Interferometer Gravitational-Wave Observatory simultaneously observed a transient gravitational-wave signal. The signal sweeps upwards in frequency from 35 to 250 Hz with a peak gravitational-wave strain of $1.0 \times 10^{-21}$. It matches the waveform predicted by general relativity for the inspiral and merger of a pair of black holes and the ringdown of the resulting single black hole. The signal was observed with a matched-filter signal-to-noise ratio of 24 and a false alarm rate estimated to be less than 1 event per 203 000 years, equivalent to a significance greater than 5.1 {\sigma}. The source lies at a luminosity distance of $410^{+160}_{-180}$ Mpc corresponding to a redshift $z = 0.09^{+0.03}_{-0.04}$. In the source frame, the initial black hole masses are $36^{+5}_{-4} M_\odot$ and $29^{+4}_{-4} M_\odot$, and the final black hole mass is $62^{+4}_{-4} M_\odot$, with $3.0^{+0.5}_{-0.5} M_\odot c^2$ radiated in gravitational waves. All uncertainties define 90% credible intervals.These observations demonstrate the existence of binary stellar-mass black hole systems. This is the first direct detection of gravitational waves and the first observation of a binary black hole merger.

8,011 citations

Proceedings ArticleDOI

21 Jul 2017-
TL;DR: This paper exploits the inherent multi-scale, pyramidal hierarchy of deep convolutional networks to construct feature pyramids with marginal extra cost and achieves state-of-the-art single-model results on the COCO detection benchmark without bells and whistles.
Abstract: Feature pyramids are a basic component in recognition systems for detecting objects at different scales. But pyramid representations have been avoided in recent object detectors that are based on deep convolutional networks, partially because they are slow to compute and memory intensive. In this paper, we exploit the inherent multi-scale, pyramidal hierarchy of deep convolutional networks to construct feature pyramids with marginal extra cost. A top-down architecture with lateral connections is developed for building high-level semantic feature maps at all scales. This architecture, called a Feature Pyramid Network (FPN), shows significant improvement as a generic feature extractor in several applications. Using a basic Faster R-CNN system, our method achieves state-of-the-art single-model results on the COCO detection benchmark without bells and whistles, surpassing all existing single-model entries including those from the COCO 2016 challenge winners. In addition, our method can run at 5 FPS on a GPU and thus is a practical and accurate solution to multi-scale object detection. Code will be made publicly available.

7,876 citations

Posted Content
TL;DR: The authors present some updates to YOLO!
Abstract: We present some updates to YOLO! We made a bunch of little design changes to make it better. We also trained this new network that's pretty swell. It's a little bigger than last time but more accurate. It's still fast though, don't worry. At 320x320 YOLOv3 runs in 22 ms at 28.2 mAP, as accurate as SSD but three times faster. When we look at the old .5 IOU mAP detection metric YOLOv3 is quite good. It achieves 57.9 mAP@50 in 51 ms on a Titan X, compared to 57.5 mAP@50 in 198 ms by RetinaNet, similar performance but 3.8x faster. As always, all the code is online at this https URL

7,812 citations

Journal ArticleDOI
18 Aug 2016-Nature
TL;DR: The aggregation and analysis of high-quality exome (protein-coding region) DNA sequence data for 60,706 individuals of diverse ancestries generated as part of the Exome Aggregation Consortium (ExAC) provides direct evidence for the presence of widespread mutational recurrence.
Abstract: Large-scale reference data sets of human genetic variation are critical for the medical and functional interpretation of DNA sequence changes. Here we describe the aggregation and analysis of high-quality exome (protein-coding region) DNA sequence data for 60,706 individuals of diverse ancestries generated as part of the Exome Aggregation Consortium (ExAC). This catalogue of human genetic diversity contains an average of one variant every eight bases of the exome, and provides direct evidence for the presence of widespread mutational recurrence. We have used this catalogue to calculate objective metrics of pathogenicity for sequence variants, and to identify genes subject to strong selection against various classes of mutation; identifying 3,230 genes with near-complete depletion of predicted protein-truncating variants, with 72% of these genes having no currently established human disease phenotype. Finally, we demonstrate that these data can be used for the efficient filtering of candidate disease-causing variants, and for the discovery of human 'knockout' variants in protein-coding genes.

7,679 citations