Abstract: The purpose of this study is to examine existing deep learning techniques for addressing class imbalanced data. Effective classification with imbalanced data is an important area of research, as high class imbalance is naturally inherent in many real-world applications, e.g., fraud detection and cancer detection. Moreover, highly imbalanced data poses added difficulty, as most learners will exhibit bias towards the majority class, and in extreme cases, may ignore the minority class altogether. Class imbalance has been studied thoroughly over the last two decades using traditional machine learning models, i.e. non-deep learning. Despite recent advances in deep learning, along with its increasing popularity, very little empirical work in the area of deep learning with class imbalance exists. Having achieved record-breaking performance results in several complex domains, investigating the use of deep neural networks for problems containing high levels of class imbalance is of great interest. Available studies regarding class imbalance and deep learning are surveyed in order to better understand the efficacy of deep learning when applied to class imbalanced data. This survey discusses the implementation details and experimental results for each study, and offers additional insight into their strengths and weaknesses. Several areas of focus include: data complexity, architectures tested, performance interpretation, ease of use, big data application, and generalization to other domains. We have found that research in this area is very limited, that most existing work focuses on computer vision tasks with convolutional neural networks, and that the effects of big data are rarely considered. Several traditional methods for class imbalance, e.g. data sampling and cost-sensitive learning, prove to be applicable in deep learning, while more advanced methods that exploit neural network feature learning abilities show promising results. The survey concludes with a discussion that highlights various gaps in deep learning from class imbalanced data for the purpose of guiding future research.

Abstract: Lithium metal is an attractive anode material for rechargeable batteries, owing to its high theoretical specific capacity of 3,860 mAh g−1. Despite extensive research efforts, there are still many fundamental challenges in using lithium metal in lithium-ion batteries. Most notably, critical information such as its nucleation and growth behaviour remains elusive. Here we explore the nucleation pattern of lithium on various metal substrates and unravel a substrate-dependent growth phenomenon that enables selective deposition of lithium metal. With the aid of binary phase diagrams, we find that no nucleation barriers are present for metals exhibiting a definite solubility in lithium, whereas appreciable nucleation barriers exist for metals with negligible solubility. We thereafter design a nanocapsule structure for lithium metal anodes consisting of hollow carbon spheres with nanoparticle seeds inside. During deposition, the lithium metal is found to predominantly grow inside the hollow carbon spheres. Such selective deposition and stable encapsulation of lithium metal eliminate dendrite formation and enable improved cycling, even in corrosive alkyl carbonate electrolytes, with 98% coulombic efficiency for more than 300 cycles. Uncontrolled lithium deposition during cycling is a major concern in the development of lithium-based batteries. Here, the authors analyse the lithium nucleation pattern on various metal substrates and demonstrate that lithium can be selectively deposited in a nanoseed inside hollow carbon spheres.

Abstract: The challenge posed by the many-body problem in quantum physics originates from the difficulty of describing the nontrivial correlations encoded in the exponential complexity of the many-body wave function. Here we demonstrate that systematic machine learning of the wave function can reduce this complexity to a tractable computational form for some notable cases of physical interest. We introduce a variational representation of quantum states based on artificial neural networks with a variable number of hidden neurons. A reinforcement-learning scheme we demonstrate is capable of both finding the ground state and describing the unitary time evolution of complex interacting quantum systems. Our approach achieves high accuracy in describing prototypical interacting spins models in one and two dimensions.

Abstract: Background Recent data suggest that complications and death from coronavirus disease 2019 (Covid-19) may be related to high viral loads. Methods In this ongoing, double-blind, phase 1–3 tr...

Abstract: Despite recent breakthroughs in the applications of deep neural networks, one setting that presents a persistent challenge is that of "one-shot learning." Traditional gradient-based networks require a lot of data to learn, often through extensive iterative training. When new data is encountered, the models must inefficiently relearn their parameters to adequately incorporate the new information without catastrophic interference. Architectures with augmented memory capacities, such as Neural Turing Machines (NTMs), offer the ability to quickly encode and retrieve new information, and hence can potentially obviate the downsides of conventional models. Here, we demonstrate the ability of a memory-augmented neural network to rapidly assimilate new data, and leverage this data to make accurate predictions after only a few samples. We also introduce a new method for accessing an external memory that focuses on memory content, unlike previous methods that additionally use memory location-based focusing mechanisms.

Abstract: Recent trends suggest that neural-network-inspired word embedding models outperform traditional count-based distributional models on word similarity and analogy detection tasks. We reveal that much of the performance gains of word embeddings are due to certain system design choices and hyperparameter optimizations, rather than the embedding algorithms themselves. Furthermore, we show that these modifications can be transferred to traditional distributional models, yielding similar gains. In contrast to prior reports, we observe mostly local or insignificant performance differences between the methods, with no global advantage to any single approach over the others.

Abstract: The superior performance of Deformable Convolutional Networks arises from its ability to adapt to the geometric variations of objects. Through an examination of its adaptive behavior, we observe that while the spatial support for its neural features conforms more closely than regular ConvNets to object structure, this support may nevertheless extend well beyond the region of interest, causing features to be influenced by irrelevant image content. To address this problem, we present a reformulation of Deformable ConvNets that improves its ability to focus on pertinent image regions, through increased modeling power and stronger training. The modeling power is enhanced through a more comprehensive integration of deformable convolution within the network, and by introducing a modulation mechanism that expands the scope of deformation modeling. To effectively harness this enriched modeling capability, we guide network training via a proposed feature mimicking scheme that helps the network to learn features that reflect the object focus and classification power of R-CNN features. With the proposed contributions, this new version of Deformable ConvNets yields significant performance gains over the original model and produces leading results on the COCO benchmark for object detection and instance segmentation.

Abstract: Organic reactions that involve the direct functionalization of non-activated C–H bonds represent an attractive class of transformations which maximize atom- and step-economy, and simplify chemical synthesis. Due to the high stability of C–H bonds, these processes, however, have most often required harsh reaction conditions, which has drastically limited their use as tools for the synthesis of complex organic molecules. Following the increased understanding of mechanistic aspects of C–H activation gained over recent years, great strides have been taken to design and develop new protocols that proceed efficiently under mild conditions and duly benefit from improved functional group tolerance and selectivity. In this review, we present the current state of the art in this field and detail C–H activation transformations reported since 2011 that proceed either at or below ambient temperature, in the absence of strongly acidic or basic additives or without strong oxidants. Furthermore, by identifying and discussing the major strategies that have led to these improvements, we hope that this review will serve as a useful conceptual overview and inspire the next generation of mild C–H transformations.

Abstract: We extensively compare, qualitatively and quantitatively, 41 state-of-the-art models (29 salient object detection, 10 fixation prediction, 1 objectness, and 1 baseline) over seven challenging data sets for the purpose of benchmarking salient object detection and segmentation methods. From the results obtained so far, our evaluation shows a consistent rapid progress over the last few years in terms of both accuracy and running time. The top contenders in this benchmark significantly outperform the models identified as the best in the previous benchmark conducted three years ago. We find that the models designed specifically for salient object detection generally work better than models in closely related areas, which in turn provides a precise definition and suggests an appropriate treatment of this problem that distinguishes it from other problems. In particular, we analyze the influences of center bias and scene complexity in model performance, which, along with the hard cases for the state-of-the-art models, provide useful hints toward constructing more challenging large-scale data sets and better saliency models. Finally, we propose probable solutions for tackling several open problems, such as evaluation scores and data set bias, which also suggest future research directions in the rapidly growing field of salient object detection.

Abstract: Recent studies on mobile network design have demonstrated the remarkable effectiveness of channel attention (e.g., the Squeeze-and-Excitation attention) for lifting model performance, but they generally neglect the positional information, which is important for generating spatially selective attention maps. In this paper, we propose a novel attention mechanism for mobile networks by embedding positional information into channel attention, which we call "coordinate attention". Unlike channel attention that transforms a feature tensor to a single feature vector via 2D global pooling, the coordinate attention factorizes channel attention into two 1D feature encoding processes that aggregate features along the two spatial directions, respectively. In this way, long-range dependencies can be captured along one spatial direction and meanwhile precise positional information can be preserved along the other spatial direction. The resulting feature maps are then encoded separately into a pair of direction-aware and position-sensitive attention maps that can be complementarily applied to the input feature map to augment the representations of the objects of interest. Our coordinate attention is simple and can be flexibly plugged into classic mobile networks, such as MobileNetV2, MobileNeXt, and EfficientNet with nearly no computational overhead. Extensive experiments demonstrate that our coordinate attention is not only beneficial to ImageNet classification but more interestingly, behaves better in down-stream tasks, such as object detection and semantic segmentation. Code is available at https://github.com/Andrew-Qibin/CoordAttention.

Abstract: The International Liaison Committee on Resuscitation (ILCOR) Advanced Life Support (ALS) Task Force performed detailed systematic reviews based on the recommendations of the Institute of Medicine of the National Academies1 and using the methodological approach proposed by the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) Working Group.2 Questions to be addressed (using the PICO [population, intervention, comparator, outcome] format)3 were prioritized by ALS Task Force members (by voting). Prioritization criteria included awareness of significant new data and new controversies or questions about practice. Questions about topics no longer relevant to contemporary practice or where little new research has occurred were given lower priority. The ALS Task Force prioritized 42 PICO questions for review. With the assistance of information specialists, a detailed search for relevant articles was performed in each of 3 online databases (PubMed, Embase, and the Cochrane Library). By using detailed inclusion and exclusion criteria, articles were screened for further evaluation. The reviewers for each question created a reconciled risk of bias assessment for each of the included studies, using state-of-the-art tools: Cochrane for randomized controlled trials (RCTs),4 Quality Assessment of Diagnostic Accuracy Studies (QUADAS)-2 for studies of diagnostic accuracy,5 and GRADE for observational studies that inform both therapy and prognosis questions.6 GRADE evidence profile tables7 were then created to facilitate an evaluation of the evidence in support of each of the critical and important outcomes. The quality of the evidence (or confidence in the estimate of the effect) was categorized as high, moderate, low, or very low,8 based on the study methodologies and the 5 core GRADE domains of risk of bias, inconsistency, indirectness, imprecision, and other considerations (including publication bias).9 These evidence profile tables were then used to create a …

Abstract: Genome-wide association studies (GWAS) have identified thousands of variants robustly associated with complex traits. However, the biological mechanisms underlying these associations are, in general, not well understood. We propose a gene-based association method called PrediXcan that directly tests the molecular mechanisms through which genetic variation affects phenotype. The approach estimates the component of gene expression determined by an individual's genetic profile and correlates 'imputed' gene expression with the phenotype under investigation to identify genes involved in the etiology of the phenotype. Genetically regulated gene expression is estimated using whole-genome tissue-dependent prediction models trained with reference transcriptome data sets. PrediXcan enjoys the benefits of gene-based approaches such as reduced multiple-testing burden and a principled approach to the design of follow-up experiments. Our results demonstrate that PrediXcan can detect known and new genes associated with disease traits and provide insights into the mechanism of these associations.

Abstract: This work presents a Multi-View Stereo system for robust and efficient dense modeling from unstructured image collections. Our core contributions are the joint estimation of depth and normal information, pixelwise view selection using photometric and geometric priors, and a multi-view geometric consistency term for the simultaneous refinement and image-based depth and normal fusion. Experiments on benchmarks and large-scale Internet photo collections demonstrate state-of-the-art performance in terms of accuracy, completeness, and efficiency.

Abstract: The efficiencies of perovskite solar cells have gone from single digits to a certified 22.1% in a few years' time. At this stage of their development, the key issues concern how to achieve further improvements in efficiency and long-term stability. We review recent developments in the quest to improve the current state of the art. Because photocurrents are near the theoretical maximum, our focus is on efforts to increase open-circuit voltage by means of improving charge-selective contacts and charge carrier lifetimes in perovskites via processes such as ion tailoring. The challenges associated with long-term perovskite solar cell device stability include the role of testing protocols, ionic movement affecting performance metrics over extended periods of time, and determination of the best ways to counteract degradation mechanisms.

Abstract: Background : Understanding the factors regulating our microbiota is important but requires appropriate statistical methodology. When comparing two or more populations most existing approaches either discount the underlying compositional structure in the microbiome data or use probability models such as the multinomial and Dirichlet-multinomial distributions, which may impose a correlation structure not suitable for microbiome data. Objective : To develop a methodology that accounts for compositional constraints to reduce false discoveries in detecting differentially abundant taxa at an ecosystem level, while maintaining high statistical power. Methods : We introduced a novel statistical framework called analysis of composition of microbiomes (ANCOM). ANCOM accounts for the underlying structure in the data and can be used for comparing the composition of microbiomes in two or more populations. ANCOM makes no distributional assumptions and can be implemented in a linear model framework to adjust for covariates as well as model longitudinal data. ANCOM also scales well to compare samples involving thousands of taxa. Results : We compared the performance of ANCOM to the standard t -test and a recently published methodology called Zero Inflated Gaussian (ZIG) methodology (1) for drawing inferences on the mean taxa abundance in two or more populations. ANCOM controlled the false discovery rate (FDR) at the desired nominal level while also improving power, whereas the t -test and ZIG had inflated FDRs, in some instances as high as 68% for the t -test and 60% for ZIG. We illustrate the performance of ANCOM using two publicly available microbial datasets in the human gut, demonstrating its general applicability to testing hypotheses about compositional differences in microbial communities. Conclusion : Accounting for compositionality using log-ratio analysis results in significantly improved inference in microbiota survey data. Keywords: constrained; relative abundance; log-ratio (Published: 29 May 2015) Citation: Microbial Ecology in Health & Disease 2015, 26: 27663 - http://dx.doi.org/10.3402/mehd.v26.27663 To access the supplementary material for this article, please see Supplementary files under ‘Article Tools’

Abstract: In this work, we propose a method for simultaneously learning features and a corresponding similarity metric for person re-identification. We present a deep convolutional architecture with layers specially designed to address the problem of re-identification. Given a pair of images as input, our network outputs a similarity value indicating whether the two input images depict the same person. Novel elements of our architecture include a layer that computes cross-input neighborhood differences, which capture local relationships between the two input images based on mid-level features from each input image. A high-level summary of the outputs of this layer is computed by a layer of patch summary features, which are then spatially integrated in subsequent layers. Our method significantly outperforms the state of the art on both a large data set (CUHK03) and a medium-sized data set (CUHK01), and is resistant to over-fitting. We also demonstrate that by initially training on an unrelated large data set before fine-tuning on a small target data set, our network can achieve results comparable to the state of the art even on a small data set (VIPeR).

Abstract: Large-scale genetic analysis of lethal phenotypes has elucidated the molecular underpinnings of many biological processes. Using the bacterial clustered regularly interspaced short palindromic repeats (CRISPR) system, we constructed a genome-wide single-guide RNA library to screen for genes required for proliferation and survival in a human cancer cell line. Our screen revealed the set of cell-essential genes, which was validated with an orthogonal gene-trap-based screen and comparison with yeast gene knockouts. This set is enriched for genes that encode components of fundamental pathways, are expressed at high levels, and contain few inactivating polymorphisms in the human population. We also uncovered a large group of uncharacterized genes involved in RNA processing, a number of whose products localize to the nucleolus. Last, screens in additional cell lines showed a high degree of overlap in gene essentiality but also revealed differences specific to each cell line and cancer type that reflect the developmental origin, oncogenic drivers, paralogous gene expression pattern, and chromosomal structure of each line. These results demonstrate the power of CRISPR-based screens and suggest a general strategy for identifying liabilities in cancer cells.

Abstract: Conditional Generative Adversarial Networks (GANs) for cross-domain image-to-image translation have made much progress recently [7, 8, 21, 12, 4, 18]. Depending on the task complexity, thousands to millions of labeled image pairs are needed to train a conditional GAN. However, human labeling is expensive, even impractical, and large quantities of data may not always be available. Inspired by dual learning from natural language translation [23], we develop a novel dual-GAN mechanism, which enables image translators to be trained from two sets of unlabeled images from two domains. In our architecture, the primal GAN learns to translate images from domain U to those in domain V, while the dual GAN learns to invert the task. The closed loop made by the primal and dual tasks allows images from either domain to be translated and then reconstructed. Hence a loss function that accounts for the reconstruction error of images can be used to train the translators. Experiments on multiple image translation tasks with unlabeled data show considerable performance gain of DualGAN over a single GAN. For some tasks, DualGAN can even achieve comparable or slightly better results than conditional GAN trained on fully labeled data.

Abstract: Importance Tumor-treating fields (TTFields) is an antimitotic treatment modality that interferes with glioblastoma cell division and organelle assembly by delivering low-intensity alternating electric fields to the tumor. Objective To investigate whether TTFields improves progression-free and overall survival of patients with glioblastoma, a fatal disease that commonly recurs at the initial tumor site or in the central nervous system. Design, Setting, and Participants In this randomized, open-label trial, 695 patients with glioblastoma whose tumor was resected or biopsied and had completed concomitant radiochemotherapy (median time from diagnosis to randomization, 3.8 months) were enrolled at 83 centers (July 2009-2014) and followed up through December 2016. A preliminary report from this trial was published in 2015; this report describes the final analysis. Interventions Patients were randomized 2:1 to TTFields plus maintenance temozolomide chemotherapy (n = 466) or temozolomide alone (n = 229). The TTFields, consisting of low-intensity, 200 kHz frequency, alternating electric fields, was delivered (≥ 18 hours/d) via 4 transducer arrays on the shaved scalp and connected to a portable device. Temozolomide was administered to both groups (150-200 mg/m2) for 5 days per 28-day cycle (6-12 cycles). Main Outcomes and Measures Progression-free survival (tested at α = .046). The secondary end point was overall survival (tested hierarchically at α = .048). Analyses were performed for the intent-to-treat population. Adverse events were compared by group. Results Of the 695 randomized patients (median age, 56 years; IQR, 48-63; 473 men [68%]), 637 (92%) completed the trial. Median progression-free survival from randomization was 6.7 months in the TTFields-temozolomide group and 4.0 months in the temozolomide-alone group (HR, 0.63; 95% CI, 0.52-0.76;P Conclusions and Relevance In the final analysis of this randomized clinical trial of patients with glioblastoma who had received standard radiochemotherapy, the addition of TTFields to maintenance temozolomide chemotherapy vs maintenance temozolomide alone, resulted in statistically significant improvement in progression-free survival and overall survival. These results are consistent with the previous interim analysis. Trial Registration clinicaltrials.gov Identifier:NCT00916409

Abstract: With a massive influx of multimodality data, the role of data analytics in health informatics has grown rapidly in the last decade. This has also prompted increasing interests in the generation of analytical, data driven models based on machine learning in health informatics. Deep learning, a technique with its foundation in artificial neural networks, is emerging in recent years as a powerful tool for machine learning, promising to reshape the future of artificial intelligence. Rapid improvements in computational power, fast data storage, and parallelization have also contributed to the rapid uptake of the technology in addition to its predictive power and ability to generate automatically optimized high-level features and semantic interpretation from the input data. This article presents a comprehensive up-to-date review of research employing deep learning in health informatics, providing a critical analysis of the relative merit, and potential pitfalls of the technique as well as its future outlook. The paper mainly focuses on key applications of deep learning in the fields of translational bioinformatics, medical imaging, pervasive sensing, medical informatics, and public health.

Abstract: Background Authors were assigned the task to develop case definitions for periodontitis in the context of the 2017 World Workshop on the Classification of Periodontal and Peri-Implant Diseases and Conditions. The aim of this manuscript is to review evidence and rationale for a revision of the current classification, to provide a framework for case definition that fully implicates state-of-the-art knowledge and can be adapted as new evidence emerges, and to suggest a case definition system that can be implemented in clinical practice, research and epidemiologic surveillance. Methods Evidence gathered in four commissioned reviews was analyzed and interpreted with special emphasis to changes with regards to the understanding available prior to the 1999 classification. Authors analyzed case definition systems employed for a variety of chronic diseases and identified key criteria for a classification/case definition of periodontitis. Results The manuscript discusses the merits of a periodontitis case definition system based on Staging and Grading and proposes a case definition framework. Stage I to IV of periodontitis is defined based on severity (primarily periodontal breakdown with reference to root length and periodontitis-associated tooth loss), complexity of management (pocket depth, infrabony defects, furcation involvement, tooth hypermobility, masticatory dysfunction) and additionally described as extent (localized or generalized). Grade of periodontitis is estimated with direct or indirect evidence of progression rate in three categories: slow, moderate and rapid progression (Grade A-C). Risk factor analysis is used as grade modifier. Conclusions The paper describes a simple matrix based on stage and grade to appropriately define periodontitis in an individual patient. The proposed case definition extends beyond description based on severity to include characterization of biological features of the disease and represents a first step towards adoption of precision medicine concepts to the management of periodontitis. It also provides the necessary framework for introduction of biomarkers in diagnosis and prognosis.

Abstract: Here, we present a major advance of the OrthoFinder method. This extends OrthoFinder’s high accuracy orthogroup inference to provide phylogenetic inference of orthologs, rooted genes trees, gene duplication events, the rooted species tree, and comparative genomic statistics. Each output is benchmarked on appropriate real or simulated datasets and, where comparable methods exist, OrthoFinder is equivalent to or outperforms these methods. Furthermore, OrthoFinder is the most accurate ortholog inference method on the Quest for Orthologs benchmark test. Finally, OrthoFinder’s comprehensive phylogenetic analysis is achieved with equivalent speed and scalability to the fastest, score-based heuristic methods. OrthoFinder is available at https://github.com/davidemms/OrthoFinder.

Abstract: We present an improved determination of the Hubble constant (H0) from Hubble Space Telescope (HST) observations of 70 long-period Cepheids in the Large Magellanic Cloud. These were obtained with the same WFC3 photometric system used to measure Cepheids in the hosts of Type Ia supernovae. Gyroscopic control of HST was employed to reduce overheads while collecting a large sample of widely-separated Cepheids. The Cepheid Period-Luminosity relation provides a zeropoint-free link with 0.4% precision between the new 1.2% geometric distance to the LMC from Detached Eclipsing Binaries (DEBs) measured by Pietrzynski et al (2019) and the luminosity of SNe Ia. Measurements and analysis of the LMC Cepheids were completed prior to knowledge of the new LMC distance. Combined with a refined calibration of the count-rate linearity of WFC3-IR with 0.1% precision (Riess et al 2019), these three improved elements together reduce the full uncertainty in the LMC geometric calibration of the Cepheid distance ladder from 2.5% to 1.3%. Using only the LMC DEBs to calibrate the ladder we find H0=74.22 +/- 1.82 km/s/Mpc including systematic uncertainties, 3% higher than before for this particular anchor. Combining the LMC DEBs, masers in NGC 4258 and Milky Way parallaxes yields our best estimate: H0 = 74.03 +/- 1.42 km/s/Mpc, including systematics, an uncertainty of 1.91%---15% lower than our best previous result. Removing any one of these anchors changes H0 by < 0.7%. The difference between H0 measured locally and the value inferred from Planck CMB+LCDM is 6.6+/-1.5 km/s/Mpc or 4.4 sigma (P=99.999% for Gaussian errors) in significance, raising the discrepancy beyond a plausible level of chance. We summarize independent tests which show this discrepancy is not readily attributable to an error in any one source or measurement, increasing the odds that it results from a cosmological feature beyond LambdaCDM.

Abstract: Adversarial examples are perturbed inputs designed to fool machine learning models. Adversarial training injects such examples into training data to increase robustness. To scale this technique to large datasets, perturbations are crafted using fast single-step methods that maximize a linear approximation of the model's loss. We show that this form of adversarial training converges to a degenerate global minimum, wherein small curvature artifacts near the data points obfuscate a linear approximation of the loss. The model thus learns to generate weak perturbations, rather than defend against strong ones. As a result, we find that adversarial training remains vulnerable to black-box attacks, where we transfer perturbations computed on undefended models, as well as to a powerful novel single-step attack that escapes the non-smooth vicinity of the input data via a small random step. We further introduce Ensemble Adversarial Training, a technique that augments training data with perturbations transferred from other models. On ImageNet, Ensemble Adversarial Training yields models with strong robustness to black-box attacks. In particular, our most robust model won the first round of the NIPS 2017 competition on Defenses against Adversarial Attacks. However, subsequent work found that more elaborate black-box attacks could significantly enhance transferability and reduce the accuracy of our models.

Abstract: We describe the first method to automatically estimate the 3D pose of the human body as well as its 3D shape from a single unconstrained image. We estimate a full 3D mesh and show that 2D joints alone carry a surprising amount of information about body shape. The problem is challenging because of the complexity of the human body, articulation, occlusion, clothing, lighting, and the inherent ambiguity in inferring 3D from 2D. To solve this, we first use a recently published CNN-based method, DeepCut, to predict (bottom-up) the 2D body joint locations. We then fit (top-down) a recently published statistical body shape model, called SMPL, to the 2D joints. We do so by minimizing an objective function that penalizes the error between the projected 3D model joints and detected 2D joints. Because SMPL captures correlations in human shape across the population, we are able to robustly fit it to very little data. We further leverage the 3D model to prevent solutions that cause interpenetration. We evaluate our method, SMPLify, on the Leeds Sports, HumanEva, and Human3.6M datasets, showing superior pose accuracy with respect to the state of the art.

Abstract: Ancient DNA research is revealing a human history far more complex than that inferred from parsimonious models based on modern DNA. Here, we review some of the key events in the peopling of the world in the light of the findings of work on ancient DNA.