3,894 research outputs found
Sparsity based defense against adversarial examples: v1.0
No full textSparsity-based defense against adversarial attacks on machine learning classifiers. Contains code for the following papers:
S. Gopalakrishnan, Z. Marzi, U. Madhow, R. Pedarsani, "Robust Adversarial Learning via Sparsifying Front Ends", arXiv:1810.10625.
Z. Marzi*, S. Gopalakrishnan*, U. Madhow, R. Pedarsani, "Sparsity-based Defense against Adversarial Attacks on Linear Classifiers", in IEEE International Symposium on Information Theory (ISIT), June 2018.
S. Gopalakrishnan*, Z. Marzi*, U. Madhow, R. Pedarsani, "Combating Adversarial Attacks Using Sparse Representations", in ICLR Workshop, April 2018
Structural health monitoring in a composite beam using magnetostrictive material through a new FE formulation
No full textA new finite element formulation was derived for structural health monitoring of laminated composite beam containing embedded magnetostrictive patches, which will act both as sensors and actuators. Coils are considered to activate patches and/or to sense the changes in stress in patches. When the actuator patch is excited dynamically by passing alternating current through the actuation coil, it introduces stress in the structure due to magneto mechanical coupling effect, which in turn produce magnetic flux in the sensing patches. This magnetic flux generates open circuit voltage in the sensing coils. Measurements of open circuit voltage before and after the damage(delamination) occurrence provides a diagnostic to indicate the presence of the damage. Numerical results have been obtained for one combination of actuator and sensor locations
Data for: Performance of Fly ash blended Crushed Sand Concrete
No full textThe dataset contains the semi-adiabatic temperature measurement of mass blocks of M50 grade concrete proportioned with (i) ordinary portland cement (OPC) with crushed sand (CS), (ii) OPC with 20% replacement of fly ash and river sand (RS) and (iii) OPC with 20% replacement and CS, designated as CCS, CFRS and CFCS respectively. This dataset also includes cube compressive strength of 200 samples of CFRS & CFCS concretes tested from different batches of field concrete, rapid chloride penetration test (RCPT) results and fines (silt) content variation of CFCS concrete used in a major construction project
Exploration of Plant Growth-Promoting Actinomycetes for Biofortification of Mineral Nutrients
No full textMineral malnutrition, especially Fe and Zn, affects more than two million people around the world and increases vulnerability to illness and infections. These malnourished people live in developing countries and rely upon staple foods routinely with inability to either afford for dietary diversification or pharmaceutical supplementation or industrial fortification of minerals. Biofortification is a strategy that can tackle hidden hunger merely through staple foods that people eat every day. This strategy can be achieved through agronomic practices and conventional breeding and genetic engineering approaches, and each has their own pros and cons. The sustainability of such grain fortification with higher seed mineral concentration is soil health dependent, especially on the availability of mineral in the rhizosphere. Microorganisms, the invisible engineers in improving the soil health by solubilizing trace elements and by driving various biogeochemical cycles of soil, have the ability to serve as a key solution for this complex issue. In specific, plant growth-promoting (PGP) microbes reside in root-soil interface and employ the use of siderophores, organic acids, and exopolysaccharides for increasing the mineral availability and subsequent mobilization to the plants. Increasing the seed mineral density with the use of such PGP microbes, especially actinomycetes, is in its infancy. Hence, this chapter is aimed to bring a view on the role of microbes, especially actinomycetes, with metal-mobilizing and PGP traits for biofortification as this strategy may act as a complementary sustainable tool for the existing biofortification strategies
Plant Growth Promoting Actinobacteria : A New Avenue for Enhancing the Productivity and Soil Fertility of Grain Legumes
No full textGlobal yields of legumes have been relatively stagnant for the last five decades, despite the adoption of conventional and molecular breeding approaches. The use of plant growth-promoting (PGP) bacteria for improving agricultural production, soil and plant health has become one of the most attractive strategies for developing sustainable agriculture. Actinomycetes are bacteria that play an important role in PGP and plant protection, produce secondary metabolites of commercial interest, and their use is well documented in wheat, rice, beans, chickpeas and peas. In order to promote legumes, the general assembly of the UN recently declared 2016 the “International Year of Pulses.” In view of this development, this book illustrates how PGP actinomycetes can improve grain yield and soil fertility, improve control of insect pests and phytopathogens, and enhance host-plant resistance. It also addresses special topics of current interest, e.g. the role of PGP actinomycetes in the biofortification of legume seeds and bioremediation of heavy metals
A report on occurrence of Phlebotomine sand flies (Diptera: Psychodidae) and two new country records from Andaman & Nicobar Islands, a Union territory of India
No full textJambulingam, P., Srinivasan, R., Gopalakrishnan, S. (2022): A report on occurrence of Phlebotomine sand flies (Diptera: Psychodidae) and two new country records from Andaman & Nicobar Islands, a Union territory of India. Zootaxa 5093 (2): 241-246, DOI: 10.11646/zootaxa.5093.2.
Understanding the Evolution of Plant Growth-Promoting Rhizobacteria
No full textSoil is an integral part of the complicated natural environment which is
very much alive with complex ecosystem of microbes. Among them, the symbiotic
association of rhizobacteria with plants especially on agriculturally important crops
is very much advantageous in improving the soil and plant health. These plant
growth-promoting rhizobacteria (PGPR) have evolved over the years and involved
in many plant functions such as growth promotion, root development, colonization,
production of metabolites and in eliciting plant defence mechanism against abiotic
and biotic agents. The PGPR’s ability to fix the atmospheric nitrogen, solubilize
phosphate, potassium and zinc, produce siderophore along with wide variety of
phytohormones and secondary metabolites such as antibiotics have attributed to their
significance as biocontrol agents. These functions lead to their application as
biofertilizers, biopesticides, bioprotectants and phytostimulators. The employment
of these PGPR is very much important in agricultural fields as they reduce the burden
of chemical fertilizers and pesticides to the farmers and in turn promises an increased
crop yield. This chapter discusses the symbiotic association of PGPR with plants in
detail including their direct and indirect mechanisms and basis of their induced
systemic defence mechanism. It also highlights the use of bioinoculants and nanoformulations
of PGPR as an effective tool towards enhanced agricultural production
and to combat the plant diseases in an eco-friendly manner
Sensing of delamination in laminated composite beams using multiple magnetostrictive patches
No full textSmart Composite laminates with multiple magnetorestrictive patches have been fabricated, one patch is for actuation and the other patch is for sensing. Horse-show type coil arrangement is used for actuation and sensing, which contains two arms, one arm is used for actuation and the other arm for sensing. Delamination sensing is carried out for composite laminated beam specimens, one with delamination and the other without any defect or damage. Actuation coil is used to induce stress in the magnetostrictive material, sensing coil is used for sensing to measure the induced open circuit voltage. Due to delamination, the stress produced in the magnetostrictive patch changes and that alters the magnetic flux through the sensing coil. This causes a change in the induced open circuit voltage across the sensing coil. The paper presents, the sensing of delamination in composite laminates using multiple magnetostrictive patches with horse-shoe type coil arrangement. The difference in the open circuit voltage with and without delamination is treated as delamination sensing voltage. For theoretical modeling, different models of the laminated composite beams with magnetostrictive sensors and actuators have been developed for the beams with and without delamination. Finite element analysis is carried out using NISA finite element software. Numerical values of open circuit voltage verses actuation current for both the experimental and computational studies are presented. Comparison of experimental and the computational studies shows that there is a good agreement between both the results obtained. Hence, the modeling method is very efficient and may be very much useful for the further studies related to the smart structures with intelligent materials
Delamination detection in aerospace composite panels using convolutional autoencoders
No full textModern aerospace structures demand lightweight design procedures and require scheduled maintenance intervals. Supervised deep learning strategies can allow reliable damage detection provided a large amount of data is available to train. These learning algorithms may face problems in the absence of possible damage scenarios in the training dataset. This class imbalance problem in supervised deep learning may curtail the learning process and can possess issues related to generalization on unseen examples. On the other hand, unsupervised deep learning algorithms like autoencoders can handle such situations in the absence of labeled data. In this study, an aerospace composite panel is interrogated with a circular array of piezoelectric transducers using ultrasonic guided waves in a round-robin fashion. The time-series signals are collected for both the healthy and unhealthy state of the structure and transformed into a time-frequency dataset using continuous wavelet transformation. A convolutional autoencoder algorithm trained on healthy signals is used to identify anomalies in the form of delamination in the structure. The proposed methodology can successfully identify delamination in the structure with good accuracy
- …
