Showing papers by "Yahoo! published in 2017"

Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations

Abstract: Despite progress in perceptual tasks such as image classification, computers still perform poorly on cognitive tasks such as image description and question answering. Cognition is core to tasks that involve not just recognizing, but reasoning about our visual world. However, models used to tackle the rich content in images for cognitive tasks are still being trained using the same datasets designed for perceptual tasks. To achieve success at cognitive tasks, models need to understand the interactions and relationships between objects in an image. When asked "What vehicle is the person riding?", computers will need to identify the objects in an image as well as the relationships riding(man, carriage) and pulling(horse, carriage) to answer correctly that "the person is riding a horse-drawn carriage." In this paper, we present the Visual Genome dataset to enable the modeling of such relationships. We collect dense annotations of objects, attributes, and relationships within each image to learn these models. Specifically, our dataset contains over 108K images where each image has an average of $$35$$35 objects, $$26$$26 attributes, and $$21$$21 pairwise relationships between objects. We canonicalize the objects, attributes, relationships, and noun phrases in region descriptions and questions answer pairs to WordNet synsets. Together, these annotations represent the densest and largest dataset of image descriptions, objects, attributes, relationships, and question answer pairs.

FeUdal Networks for Hierarchical Reinforcement Learning

Abstract: We introduce FeUdal Networks (FuNs): a novel architecture for hierarchical reinforcement learning. Our approach is inspired by the feudal reinforcement learning proposal of Dayan and Hinton, and gains power and efficacy by decoupling end-to-end learning across multiple levels -- allowing it to utilise different resolutions of time. Our framework employs a Manager module and a Worker module. The Manager operates at a lower temporal resolution and sets abstract goals which are conveyed to and enacted by the Worker. The Worker generates primitive actions at every tick of the environment. The decoupled structure of FuN conveys several benefits -- in addition to facilitating very long timescale credit assignment it also encourages the emergence of sub-policies associated with different goals set by the Manager. These properties allow FuN to dramatically outperform a strong baseline agent on tasks that involve long-term credit assignment or memorisation. We demonstrate the performance of our proposed system on a range of tasks from the ATARI suite and also from a 3D DeepMind Lab environment.

Learning from Noisy Labels with Distillation

Abstract: The ability of learning from noisy labels is very useful in many visual recognition tasks, as a vast amount of data with noisy labels are relatively easy to obtain. Traditionally, label noise has been treated as statistical outliers, and techniques such as importance re-weighting and bootstrapping have been proposed to alleviate the problem. According to our observation, the real-world noisy labels exhibit multimode characteristics as the true labels, rather than behaving like independent random outliers. In this work, we propose a unified distillation framework to use “side” information, including a small clean dataset and label relations in knowledge graph, to “hedge the risk” of learning from noisy labels. Unlike the traditional approaches evaluated based on simulated label noises, we propose a suite of new benchmark datasets, in Sports, Species and Artifacts domains, to evaluate the task of learning from noisy labels in the practical setting. The empirical study demonstrates the effectiveness of our proposed method in all the domains.

Biological Activities of Essential Oils: From Plant Chemoecology to Traditional Healing Systems

Abstract: Essential oils are complex mixtures of hydrocarbons and their oxygenated derivatives arising from two different isoprenoid pathways. Essential oils are produced by glandular trichomes and other secretory structures, specialized secretory tissues mainly diffused onto the surface of plant organs, particularly flowers and leaves, thus exerting a pivotal ecological role in plant. In addition, essential oils have been used, since ancient times, in many different traditional healing systems all over the world, because of their biological activities. Many preclinical studies have documented antimicrobial, antioxidant, anti-inflammatory and anticancer activities of essential oils in a number of cell and animal models, also elucidating their mechanism of action and pharmacological targets, though the paucity of in human studies limits the potential of essential oils as effective and safe phytotherapeutic agents. More well-designed clinical trials are needed in order to ascertain the real efficacy and safety of these plant products.

Embedding-based News Recommendation for Millions of Users

Abstract: It is necessary to understand the content of articles and user preferences to make effective news recommendations. While ID-based methods, such as collaborative filtering and low-rank factorization, are well known for making recommendations, they are not suitable for news recommendations because candidate articles expire quickly and are replaced with new ones within short spans of time. Word-based methods, which are often used in information retrieval settings, are good candidates in terms of system performance but have issues such as their ability to cope with synonyms and orthographical variants and define "queries" from users' historical activities. This paper proposes an embedding-based method to use distributed representations in a three step end-to-end manner: (i) start with distributed representations of articles based on a variant of a denoising autoencoder, (ii) generate user representations by using a recurrent neural network (RNN) with browsing histories as input sequences, and (iii) match and list articles for users based on inner-product operations by taking system performance into consideration. The proposed method performed well in an experimental offline evaluation using past access data on Yahoo! JAPAN's homepage. We implemented it on our actual news distribution system based on these experimental results and compared its online performance with a method that was conventionally incorporated into the system. As a result, the click-through rate (CTR) improved by 23% and the total duration improved by 10%, compared with the conventionally incorporated method. Services that incorporated the method we propose are already open to all users and provide recommendations to over ten million individual users per day who make billions of accesses per month.

Deep and Confident Prediction for Time Series at Uber

Abstract: Reliable uncertainty estimation for time series prediction is critical in many fields, including physics, biology, and manufacturing. At Uber, probabilistic time series forecasting is used for robust prediction of number of trips during special events, driver incentive allocation, as well as real-time anomaly detection across millions of metrics. Classical time series models are often used in conjunction with a probabilistic formulation for uncertainty estimation. However, such models are hard to tune, scale, and add exogenous variables to. Motivated by the recent resurgence of Long Short Term Memory networks, we propose a novel end-to-end Bayesian deep model that provides time series prediction along with uncertainty estimation. We provide detailed experiments of the proposed solution on completed trips data, and successfully apply it to large-scale time series anomaly detection at Uber.

The Role of Tumor Microenvironment in Chemoresistance: To Survive, Keep Your Enemies Closer

Abstract: Chemoresistance is a leading cause of morbidity and mortality in cancer and it continues to be a challenge in cancer treatment. Chemoresistance is influenced by genetic and epigenetic alterations which affect drug uptake, metabolism and export of drugs at the cellular levels. While most research has focused on tumor cell autonomous mechanisms of chemoresistance, the tumor microenvironment has emerged as a key player in the development of chemoresistance and in malignant progression, thereby influencing the development of novel therapies in clinical oncology. It is not surprising that the study of the tumor microenvironment is now considered to be as important as the study of tumor cells. Recent advances in technological and analytical methods, especially ‘omics’ technologies, has made it possible to identify specific targets in tumor cells and within the tumor microenvironment to eradicate cancer. Tumors need constant support from previously ‘unsupportive’ microenvironments. Novel therapeutic strategies that inhibit such microenvironmental support to tumor cells would reduce chemoresistance and tumor relapse. Such strategies can target stromal cells, proteins released by stromal cells and non-cellular components such as the extracellular matrix (ECM) within the tumor microenvironment. Novel in vitro tumor biology models that recapitulate the in vivo tumor microenvironment such as multicellular tumor spheroids, biomimetic scaffolds and tumor organoids are being developed and are increasing our understanding of cancer cell-microenvironment interactions. This review offers an analysis of recent developments on the role of the tumor microenvironment in the development of chemoresistance and the strategies to overcome microenvironment-mediated chemoresistance. We propose a systematic analysis of the relationship between tumor cells and their respective tumor microenvironments and our data show that, to survive, cancer cells interact closely with tumor microenvironment components such as mesenchymal stem cells and the extracellular matrix.

FeUdal Networks for Hierarchical Reinforcement Learning

Abstract: We introduce FeUdal Networks (FuNs): a novel architecture for hierarchical reinforcement learning. Our approach is inspired by the feudal reinforcement learning proposal of Dayan and Hinton, and gains power and efficacy by decoupling end-to-end learning across multiple levels -- allowing it to utilise different resolutions of time. Our framework employs a Manager module and a Worker module. The Manager operates at a lower temporal resolution and sets abstract goals which are conveyed to and enacted by the Worker. The Worker generates primitive actions at every tick of the environment. The decoupled structure of FuN conveys several benefits -- in addition to facilitating very long timescale credit assignment it also encourages the emergence of sub-policies associated with different goals set by the Manager. These properties allow FuN to dramatically outperform a strong baseline agent on tasks that involve long-term credit assignment or memorisation. We demonstrate the performance of our proposed system on a range of tasks from the ATARI suite and also from a 3D DeepMind Lab environment.

TGIF-QA: Toward Spatio-Temporal Reasoning in Visual Question Answering

Abstract: Vision and language understanding has emerged as a subject undergoing intense study in Artificial Intelligence. Among many tasks in this line of research, visual question answering (VQA) has been one of the most successful ones, where the goal is to learn a model that understands visual content at region-level details and finds their associations with pairs of questions and answers in the natural language form. Despite the rapid progress in the past few years, most existing work in VQA have focused primarily on images. In this paper, we focus on extending VQA to the video domain and contribute to the literature in three important ways. First, we propose three new tasks designed specifically for video VQA, which require spatio-temporal reasoning from videos to answer questions correctly. Next, we introduce a new large-scale dataset for video VQA named TGIF-QA that extends existing VQA work with our new tasks. Finally, we propose a dual-LSTM based approach with both spatial and temporal attention, and show its effectiveness over conventional VQA techniques through empirical evaluations.

Learning from Noisy Labels with Distillation

Abstract: The ability of learning from noisy labels is very useful in many visual recognition tasks, as a vast amount of data with noisy labels are relatively easy to obtain. Traditionally, the label noises have been treated as statistical outliers, and approaches such as importance re-weighting and bootstrap have been proposed to alleviate the problem. According to our observation, the real-world noisy labels exhibit multi-mode characteristics as the true labels, rather than behaving like independent random outliers. In this work, we propose a unified distillation framework to use side information, including a small clean dataset and label relations in knowledge graph, to "hedge the risk" of learning from noisy labels. Furthermore, unlike the traditional approaches evaluated based on simulated label noises, we propose a suite of new benchmark datasets, in Sports, Species and Artifacts domains, to evaluate the task of learning from noisy labels in the practical setting. The empirical study demonstrates the effectiveness of our proposed method in all the domains.

Jointly Attentive Spatial-Temporal Pooling Networks for Video-Based Person Re-identification

Abstract: Person Re-Identification (person re-id) is a crucial task as its applications in visual surveillance and human-computer interaction. In this work, we present a novel joint Spatial and Temporal Attention Pooling Network (ASTPN) for video-based person re-identification, which enables the feature extractor to be aware of the current input video sequences, in a way that interdependency from the matching items can directly influence the computation of each other's representation. Specifically, the spatial pooling layer is able to select regions from each frame, while the attention temporal pooling performed can select informative frames over the sequence, both pooling guided by the information from distance matching. Experiments are conduced on the iLIDS-VID, PRID-2011 and MARS datasets and the results demonstrate that this approach outperforms existing state-of-art methods. We also analyze how the joint pooling in both dimensions can boost the person re-id performance more effectively than using either of them separately 1.

The Bitcoin Backbone Protocol with Chains of Variable Difficulty

Abstract: Bitcoin’s innovative and distributedly maintained blockchain data structure hinges on the adequate degree of difficulty of so-called “proofs of work,” which miners have to produce in order for transactions to be inserted. Importantly, these proofs of work have to be hard enough so that miners have an opportunity to unify their views in the presence of an adversary who interferes but has bounded computational power, but easy enough to be solvable regularly and enable the miners to make progress. As such, as the miners’ population evolves over time, so should the difficulty of these proofs. Bitcoin provides this adjustment mechanism, with empirical evidence of a constant block generation rate against such population changes.

Aging Gut Microbiota at the Cross-Road between Nutrition, Physical Frailty, and Sarcopenia: Is There a Gut-Muscle Axis?

Abstract: Inadequate nutrition and physical inactivity are the mainstays of primary sarcopenia–physiopathology in older individuals. Gut microbiota composition is strongly dependent on both of these elements, and conversely, can also influence the host physiology by modulating systemic inflammation, anabolism, insulin sensitivity, and energy production. The bacterial metabolism of nutrients theoretically influences skeletal muscle cell functionality through producing mediators that drive all of these systemic effects. In this study, we review the scientific literature supporting the concept of the involvement of gut microbiota in primary sarcopenia physiopathology. First, we examine studies associating fecal microbiota alterations with physical frailty, i.e., the loss of muscle performance and normal muscle mass. Then, we consider studies exploring the effects of exercise on gut microbiota composition. Finally, we examine studies demonstrating the possible effects of mediators produced by gut microbiota on skeletal muscle, and intervention studies considering the effects of prebiotic or probiotic administration on muscle function. Even if there is no evidence of a distinct gut microbiota composition in older sarcopenic patients, we conclude that the literature supports the possible presence of a “gut–muscle axis”, whereby gut microbiota may act as the mediator of the effects of nutrition on muscle cells.

Mining Fashion Outfit Composition Using an End-to-End Deep Learning Approach on Set Data

Abstract: Composing fashion outfits involves deep under-standing of fashion standards while incorporating creativity for choosing multiple fashion items (e.g., jewelry, bag, pants, dress). In fashion websites, popular or high-quality fashion outfits are usually designed by fashion experts and followed by large audiences. In this paper, we propose a machine learning system to compose fashion outfits automatically. The core of the proposed automatic composition system is to score fashion outfit candidates based on the appearances and metadata. We propose to leverage outfit popularity on fashion-oriented websites to supervise the scoring component. The scoring component is a multimodal multiinstance deep learning system that evaluates instance aesthetics and set compatibility simultaneously. In order to train and evaluate the proposed composition system, we have collected a large-scale fashion outfit dataset with 195K outfits and 368K fashion items from Polyvore. Although the fashion outfit scoring and composition is rather challenging, we have achieved an AUC of 85% for the scoring component, and an accuracy of 77% for a constrained composition task.

Interpretable Predictions of Tree-based Ensembles via Actionable Feature Tweaking

Abstract: Machine-learned models are often described as "black boxes". In many real-world applications however, models may have to sacrifice predictive power in favour of human-interpretability. When this is the case, feature engineering becomes a crucial task, which requires significant and time-consuming human effort. Whilst some features are inherently static, representing properties that cannot be influenced (e.g., the age of an individual), others capture characteristics that could be adjusted (e.g., the daily amount of carbohydrates taken). Nonetheless, once a model is learned from the data, each prediction it makes on new instances is irreversible - assuming every instance to be a static point located in the chosen feature space. There are many circumstances however where it is important to understand (i) why a model outputs a certain prediction on a given instance, (ii) which adjustable features of that instance should be modified, and finally (iii) how to alter such a prediction when the mutated instance is input back to the model. In this paper, we present a technique that exploits the internals of a tree-based ensemble classifier to offer recommendations for transforming true negative instances into positively predicted ones. We demonstrate the validity of our approach using an online advertising application. First, we design a Random Forest classifier that effectively separates between two types of ads: low (negative) and high (positive) quality ads (instances). Then, we introduce an algorithm that provides recommendations that aim to transform a low quality ad (negative instance) into a high quality one (positive instance). Finally, we evaluate our approach on a subset of the active inventory of a large ad network, Yahoo Gemini.

JFLEG: A Fluency Corpus and Benchmark for Grammatical Error Correction

Abstract: We present a new parallel corpus, JHU FLuency-Extended GUG corpus (JFLEG) for developing and evaluating grammatical error correction (GEC). Unlike other corpora, it represents a broad range of language proficiency levels and uses holistic fluency edits to not only correct grammatical errors but also make the original text more native sounding. We describe the types of corrections made and benchmark four leading GEC systems on this corpus, identifying specific areas in which they do well and how they can improve. JFLEG fulfills the need for a new gold standard to properly assess the current state of GEC.

Improving Pairwise Ranking for Multi-label Image Classification

Abstract: Learning to rank has recently emerged as an attractive technique to train deep convolutional neural networks for various computer vision tasks. Pairwise ranking, in particular, has been successful in multi-label image classification, achieving state-of-the-art results on various benchmarks. However, most existing approaches use the hinge loss to train their models, which is non-smooth and thus is difficult to optimize especially with deep networks. Furthermore, they employ simple heuristics, such as top-k or thresholding, to determine which labels to include in the output from a ranked list of labels, which limits their use in the real-world setting. In this work, we propose two techniques to improve pairwise ranking based multi-label image classification by solving the aforementioned problems: (1) we propose a novel loss function for pairwise ranking, which is smooth everywhere, and (2) we incorporate a label decision module into the model, estimating the optimal confidence thresholds for each visual concept. We provide theoretical analyses of our loss function from the point of view of the Bayes consistency and risk minimization, and show its benefit over existing pairwise ranking formulations. We also demonstrate the effectiveness of our approach on two large-scale datasets, NUS-WIDE and MS-COCO, achieving the best reported result in the literature.

Improving the Care and Treatment of Monkeypox Patients in Low-Resource Settings: Applying Evidence from Contemporary Biomedical and Smallpox Biodefense Research

Abstract: Monkeypox is a smallpox-like illness that can be accompanied by a range of significant medical complications. To date there are no standard or optimized guidelines for the clinical management of monkeypox (MPX) patients, particularly in low-resource settings. Consequently, patients can experience protracted illness and poor outcomes. Improving care necessitates developing a better understanding of the range of clinical manifestations-including complications and sequelae-as well as of features of illness that may be predictive of illness severity and poor outcomes. Experimental and natural infection of non-human primates with monkeypox virus can inform the approach to improving patient care, and may suggest options for pharmaceutical intervention. These studies have traditionally been performed to address the threat of smallpox bioterrorism and were designed with the intent of using MPX as a disease surrogate for smallpox. In many cases this necessitated employing high-dose, inhalational or intravenous challenge to recapitulate the severe manifestations of illness seen with smallpox. Overall, these data-and data from biomedical research involving burns, superficial wounds, herpes, eczema vaccinatum, and so forth-suggest that MPX patients could benefit from clinical support to mitigate the consequences of compromised skin and mucosa. This should include prevention and treatment of secondary bacterial infections (and other complications), ensuring adequate hydration and nutrition, and protecting vulnerable anatomical locations such as the eyes and genitals. A standard of care that considers these factors should be developed and assessed in different settings, using clinical metrics specific for MPX alongside consideration of antiviral therapies.

Nuts and Human Health Outcomes: A Systematic Review.

Abstract: There has been increasing interest in nuts and their outcome regarding human health. The consumption of nuts is frequently associated with reduction in risk factors for chronic diseases. Although nuts are high calorie foods, several studies have reported beneficial effects after nut consumption, due to fatty acid profiles, vegetable proteins, fibers, vitamins, minerals, carotenoids, and phytosterols with potential antioxidant action. However, the current findings about the benefits of nut consumption on human health have not yet been clearly discussed. This review highlights the effects of nut consumption on the context of human health.

Improving Bread Quality with the Application of a Newly Purified Thermostable α-Amylase from Rhizopus oryzae FSIS4

Abstract: A new thermostable α-amylase from Rhizopus oryzae FSIS4 was purified for first time and recovered in a single step using a three-phase partitioning (TPP) system. The fungal α-amylase, at a concentration of 1.936 U per kg of flour, was used in bread-making and compared to the commercial enzyme. The results showed a significant effect of the recovered α-amylase in the prepared bread and allowed us to improve the quality of the bread. The study indicated clearly that the recovered α-amylase is a potential candidate for future applications in the bread-making industry and in other food biotechnology applications.

TGIF-QA: Toward Spatio-Temporal Reasoning in Visual Question Answering

Abstract: Vision and language understanding has emerged as a subject undergoing intense study in Artificial Intelligence. Among many tasks in this line of research, visual question answering (VQA) has been one of the most successful ones, where the goal is to learn a model that understands visual content at region-level details and finds their associations with pairs of questions and answers in the natural language form. Despite the rapid progress in the past few years, most existing work in VQA have focused primarily on images. In this paper, we focus on extending VQA to the video domain and contribute to the literature in three important ways. First, we propose three new tasks designed specifically for video VQA, which require spatio-temporal reasoning from videos to answer questions correctly. Next, we introduce a new large-scale dataset for video VQA named TGIF-QA that extends existing VQA work with our new tasks. Finally, we propose a dual-LSTM based approach with both spatial and temporal attention, and show its effectiveness over conventional VQA techniques through empirical evaluations.

Improving Pairwise Ranking for Multi-label Image Classification

Abstract: Learning to rank has recently emerged as an attractive technique to train deep convolutional neural networks for various computer vision tasks. Pairwise ranking, in particular, has been successful in multi-label image classification, achieving state-of-the-art results on various benchmarks. However, most existing approaches use the hinge loss to train their models, which is non-smooth and thus is difficult to optimize especially with deep networks. Furthermore, they employ simple heuristics, such as top-k or thresholding, to determine which labels to include in the output from a ranked list of labels, which limits their use in the real-world setting. In this work, we propose two techniques to improve pairwise ranking based multi-label image classification: (1) we propose a novel loss function for pairwise ranking, which is smooth everywhere and thus is easier to optimize; and (2) we incorporate a label decision module into the model, estimating the optimal confidence thresholds for each visual concept. We provide theoretical analyses of our loss function in the Bayes consistency and risk minimization framework, and show its benefit over existing pairwise ranking formulations. We demonstrate the effectiveness of our approach on three large-scale datasets, VOC2007, NUS-WIDE and MS-COCO, achieving the best reported results in the literature.

Limited individual attention and online virality of low-quality information

Abstract: Why does low-quality information go viral? A stylized model of social media predicts that under real-world conditions of high information load and limited attention, low- and high-quality information are equally likely to go viral.

Morphological and Structural Analysis of Polyaniline and Poly(o-anisidine) Layers Generated in a DC Glow Discharge Plasma by Using an Oblique Angle Electrode Deposition Configuration

Abstract: This work is focused on the structural and morphological investigations of polyaniline and poly(o-anisidine) polymers generated in a direct current glow discharge plasma, in the vapors of the monomers, without a buffer gas, using an oblique angle-positioned substrate configuration. By atomic force microscopy and scanning electron microscopy we identified the formation of worm-like interlinked structures on the surface of the polyaniline layers, the layers being compact in the bulk. The poly(o-anisidine) layers are flat with no kind of structures on their surfaces. By Fourier transform infrared spectroscopy we identified the main IR bands characteristic of polyaniline and poly(o-anisidine), confirming that the polyaniline chemical structure is in the emeraldine form. The IR band from 1070 cm-1 was attributed to the emeraldine salt form of polyaniline as an indication of its doping with H⁺. The appearance of the IR band at 1155 cm-1 also indicates the conducting protonated of polyaniline. The X-ray diffraction revealed the formation of crystalline domains embedded in an amorphous matrix within the polyaniline layers. The interchain separation length of 3.59 A is also an indicator of the conductive character of the polymers. The X-ray diffraction pattern of poly(o-anisidine) highlights the semi-crystalline nature of the layers. The electrical conductivities of polyaniline and poly(o-anisidine) layers and their dependence with temperature are also investigated.

Sobolev Training for Neural Networks

Abstract: At the heart of deep learning we aim to use neural networks as function approximators - training them to produce outputs from inputs in emulation of a ground truth function or data creation process. In many cases we only have access to input-output pairs from the ground truth, however it is becoming more common to have access to derivatives of the target output with respect to the input -- for example when the ground truth function is itself a neural network such as in network compression or distillation. Generally these target derivatives are not computed, or are ignored. This paper introduces Sobolev Training for neural networks, which is a method for incorporating these target derivatives in addition the to target values while training. By optimising neural networks to not only approximate the function’s outputs but also the function’s derivatives we encode additional information about the target function within the parameters of the neural network. Thereby we can improve the quality of our predictors, as well as the data-efficiency and generalization capabilities of our learned function approximation. We provide theoretical justifications for such an approach as well as examples of empirical evidence on three distinct domains: regression on classical optimisation datasets, distilling policies of an agent playing Atari, and on large-scale applications of synthetic gradients. In all three domains the use of Sobolev Training, employing target derivatives in addition to target values, results in models with higher accuracy and stronger generalisation.

Blue and Red LED Illumination Improves Growth and Bioactive Compounds Contents in Acyanic and Cyanic Ocimum basilicum L. Microgreens.

Abstract: Microgreens are an excellent source of health-maintaining compounds, and the accumulation of these compounds in plant tissues may be stimulated by exogenous stimuli. While light quality effects on green basil microgreens are known, the present paper aims at improving the quality of acyanic (green) and cyanic (red) basil microgreens with different ratios of LED blue and red illumination. Growth, assimilatory and anthocyanin pigments, chlorophyll fluorescence, total phenolic, flavonoids, selected phenolic acid contents and antioxidant activity were assessed in microgreens grown for 17 days. Growth of microgreens was enhanced with predominantly blue illumination, larger cotyledon area and higher fresh mass. The same treatment elevated chlorophyll a and anthocyanin pigments contents. Colored light treatments decreased chlorophyll fluorescence ΦPSII values significantly in the green cultivar. Stimulation of phenolic synthesis and free radical scavenging activity were improved by predominantly red light in the green cultivar (up to 1.87 fold) and by predominantly blue light in the red cultivar (up to 1.73 fold). Rosmarinic and gallic acid synthesis was higher (up to 15- and 4-fold, respectively, compared to white treatment) in predominantly blue illumination. Red and blue LED ratios can be tailored to induce superior growth and phenolic contents in both red and green basil microgreens, as a convenient tool for producing higher quality foods.

Biofilms in Endodontics-Current Status and Future Directions.

Abstract: Microbiota are found in highly organized and complex entities, known as biofilms, the characteristics of which are fundamentally different from microbes in planktonic suspensions. Root canal infections are biofilm mediated. The complexity and variability of the root canal system, together with the multi-species nature of biofilms, make disinfection of this system extremely challenging. Microbial persistence appears to be the most important factor for failure of root canal treatment and this could further have an impact on pain and quality of life. Biofilm removal is accomplished by a chemo-mechanical process, using specific instruments and disinfecting chemicals in the form of irrigants and/or intracanal medicaments. Endodontic research has focused on the characterization of root canal biofilms and the clinical methods to disrupt the biofilms in addition to achieving microbial killing. In this narrative review, we discuss the role of microbial biofilms in endodontics and review the literature on the role of root canal disinfectants and disinfectant-activating methods on biofilm removal.

Diet, Gut Microbiome and Epigenetics: Emerging Links with Inflammatory Bowel Diseases and Prospects for Management and Prevention

Abstract: Inflammatory bowel diseases (IBD) represent a growing public health concern due to increasing incidence worldwide. The current notion on the pathogenesis of IBD is that genetically susceptible individuals develop intolerance to dysregulated gut microflora (dysbiosis) and chronic inflammation develops as a result of environmental triggers. Among the environmental factors associated with IBD, diet plays an important role in modulating the gut microbiome, influencing epigenetic changes, and, therefore, could be applied as a therapeutic tool to improve the disease course. Nevertheless, the current dietary recommendations for disease prevention and management are scarce and have weak evidence. This review summarises the current knowledge on the complex interactions between diet, microbiome and epigenetics in IBD. Whereas an overabundance of calories and some macronutrients increase gut inflammation, several micronutrients have the potential to modulate it. Immunonutrition has emerged as a new concept putting forward the importance of vitamins such as vitamins A, C, E, and D, folic acid, beta carotene and trace elements such as zinc, selenium, manganese and iron. However, when assessed in clinical trials, specific micronutrients exerted a limited benefit. Beyond nutrients, an anti-inflammatory dietary pattern as a complex intervention approach has become popular in recent years. Hence, exclusive enteral nutrition in paediatric Crohn’s disease is the only nutritional intervention currently recommended as a first-line therapy. Other nutritional interventions or specific diets including the Specific Carbohydrate Diet (SCD), the low fermentable oligosaccharides, disaccharides, monosaccharides, and polyol (FODMAP) diet and, most recently, the Mediterranean diet have shown strong anti-inflammatory properties and show promise for improving disease symptoms. More work is required to evaluate the role of individual food compounds and complex nutritional interventions with the potential to decrease inflammation as a means of prevention and management of IBD.

Recent Advances in 3D Printing of Aliphatic Polyesters

Abstract: 3D printing represents a valuable alternative to traditional processing methods, clearly demonstrated by the promising results obtained in the manufacture of various products, such as scaffolds for regenerative medicine, artificial tissues and organs, electronics, components for the automotive industry, art objects and so on. This revolutionary technique showed unique capabilities for fabricating complex structures, with precisely controlled physical characteristics, facile tunable mechanical properties, biological functionality and easily customizable architecture. In this paper, we provide an overview of the main 3D-printing technologies currently employed in the case of poly (lactic acid) (PLA) and polyhydroxyalkanoates (PHA), two of the most important classes of thermoplastic aliphatic polyesters. Moreover, a short presentation of the main 3D-printing methods is briefly discussed. Both PLA and PHA, in the form of filaments or powder, proved to be suitable for the fabrication of artificial tissue or scaffolds for bone regeneration. The processability of PLA and PHB blends and composites fabricated through different 3D-printing techniques, their final characteristics and targeted applications in bioengineering are thoroughly reviewed.

Malware Detection in Adversarial Settings: Exploiting Feature Evolutions and Confusions in Android Apps

Abstract: Existing techniques on adversarial malware generation employ feature mutations based on feature vectors extracted from malware. However, most (if not all) of these techniques suffer from a common limitation: feasibility of these attacks is unknown. The synthesized mutations may break the inherent constraints posed by code structures of the malware, causing either crashes or malfunctioning of malicious payloads. To address the limitation, we present Malware Recomposition Variation (MRV), an approach that conducts semantic analysis of existing malware to systematically construct new malware variants for malware detectors to test and strengthen their detection signatures/models. In particular, we use two variation strategies (i.e., malware evolution attack and malware confusion attack) following structures of existing malware to enhance feasibility of the attacks. Upon the given malware, we conduct semantic-feature mutation analysis and phylogenetic analysis to synthesize mutation strategies. Based on these strategies, we perform program transplantation to automatically mutate malware bytecode to generate new malware variants. We evaluate our MRV approach on actual malware variants, and our empirical evaluation on 1,935 Android benign apps and 1,917 malware shows that MRV produces malware variants that can have high likelihood to evade detection while still retaining their malicious behaviors. We also propose and evaluate three defense mechanisms to counter MRV.