arXiv subjects

Akshay Agarwal

Publications and source records attributed to Akshay Agarwal.

30 records · Page 2Linked to original sources

arXiv2020

Attack Agnostic Adversarial Defense via Visual Imperceptible Bound

The high susceptibility of deep learning algorithms against structured and unstructured perturbations has motivated the development of efficient adversarial defense algorithms. However, the lack of generalizability of existing defense algorithms and the high variability in the performance of the attack algorithms for different databases raises several questions on the effectiveness of the defense algorithms. In this research, we aim to design a defense model that is robust within a certain bound against both seen and unseen adversarial attacks. This bound is related to the visual appearance of an image, and we termed it as \textit{Visual Imperceptible Bound (VIB)}. To compute this bound, we propose a novel method that uses the database characteristics. The VIB is further used to measure the effectiveness of attack algorithms. The performance of the proposed defense model is evaluated on the MNIST, CIFAR-10, and Tiny ImageNet databases on multiple attacks that include C\&W ($l_2$) and DeepFool. The proposed defense model is not only able to increase the robustness against several attacks but also retain or improve the classification accuracy on an original clean test set. The proposed algorithm is attack agnostic, i.e. it does not require any knowledge of the attack algorithm.

cs.CV↗

arXiv2020

Image-Histogram-based Secondary Electron Counting to Evaluate Detective Quantum Efficiency in SEM

Scanning electron microscopy is a powerful tool for nanoscale imaging of organic and inorganic materials. An important metric for characterizing the limits of performance of these microscopes is the Detective Quantum Efficiency (DQE), which measures the fraction of emitted secondary electrons (SEs) that are detected by the SE detector. However, common techniques for measuring DQE approximate the SE emission process to be Poisson distributed, which can lead to incorrect DQE values. In this paper, we introduce a technique for measuring DQE in which we directly count the mean number of secondary electrons detected from a sample using image histograms. This technique does not assume Poisson distribution of SEs and makes it possible to accurately measure DQE for a wider range of imaging conditions. As a demonstration of our technique, we map the variation of DQE as a function of working distance in the microscope.

physics.ins-det↗

arXiv2020

Securing CNN Model and Biometric Template using Blockchain

Blockchain has emerged as a leading technology that ensures security in a distributed framework. Recently, it has been shown that blockchain can be used to convert traditional blocks of any deep learning models into secure systems. In this research, we model a trained biometric recognition system in an architecture which leverages the blockchain technology to provide fault tolerant access in a distributed environment. The advantage of the proposed approach is that tampering in one particular component alerts the whole system and helps in easy identification of `any' possible alteration. Experimentally, with different biometric modalities, we have shown that the proposed approach provides security to both deep learning model and the biometric template.

cs.CR↗

arXiv2020

Large-area microwire MoSi single-photon detectors at 1550 nm wavelength

We demonstrate saturated internal detection efficiency at 1550 nm wavelengths for meander-shaped superconducting nanowire single-photon detectors made of 3 nm thick MoSi films with widths of 1 and 3 $μm$, and active areas up to 400 by 400 $μm^2$. Despite hairpin turns and a large number of squares (up to $10^4$) in the device, the dark count rate was measured to be ~10$^3$ cps at 99% of the switching current. This value is about two orders of magnitude lower than results reported recently for short MoSi devices with shunt resistors. We also found that 5 nm thick MoSi detectors with the same geometry were insensitive to single near-infrared photons, which may be associated with different levels of suppression of the superconducting order parameter. However, our results obtained on 3 nm thick MoSi devices are in a good agreement with predictions in the frame of a kinetic-equation approach.

physics.app-ph↗

arXiv2020

IBM Functional Genomics Platform, A Cloud-Based Platform for Studying Microbial Life at Scale

The rapid growth in biological sequence data is revolutionizing our understanding of genotypic diversity and challenging conventional approaches to informatics. With the increasing availability of genomic data, traditional bioinformatic tools require substantial computational time and the creation of ever-larger indices each time a researcher seeks to gain insight from the data. To address these challenges, we pre-computed important relationships between biological entities spanning the Central Dogma of Molecular Biology and captured this information in a relational database. The database can be queried across hundreds of millions of entities and returns results in a fraction of the time required by traditional methods. In this paper, we describe \textit{IBM Functional Genomics Platform} (formerly known as OMXWare), a comprehensive database relating genotype to phenotype for bacterial life. Continually updated, IBM Functional Genomics Platform today contains data derived from 200,000 curated, self-consistently assembled genomes. The database stores functional data for over 68 million genes, 52 million proteins, and 239 million domains with associated biological activity annotations from Gene Ontology, KEGG, MetaCyc, and Reactome. IBM Functional Genomics Platform maps all of the many-to-many connections between each biological entity including the originating genome, gene, protein, and protein domain. Various microbial studies, from infectious disease to environmental health, can benefit from the rich data and connections. We describe the data selection, the pipeline to create and update the IBM Functional Genomics Platform, and the developer tools (Python SDK and REST APIs) which allow researchers to efficiently study microbial life at scale.

q-bio.QM↗

arXiv2020

On the Robustness of Face Recognition Algorithms Against Attacks and Bias

Face recognition algorithms have demonstrated very high recognition performance, suggesting suitability for real world applications. Despite the enhanced accuracies, robustness of these algorithms against attacks and bias has been challenged. This paper summarizes different ways in which the robustness of a face recognition algorithm is challenged, which can severely affect its intended working. Different types of attacks such as physical presentation attacks, disguise/makeup, digital adversarial attacks, and morphing/tampering using GANs have been discussed. We also present a discussion on the effect of bias on face recognition models and showcase that factors such as age and gender variations affect the performance of modern algorithms. The paper also presents the potential reasons for these challenges and some of the future research directions for increasing the robustness of face recognition models.

cs.CV↗

arXiv2019

Reduced damage in electron microscopy by using interaction-free measurement and conditional re-illumination

Interaction-free measurement (IFM) has been proposed as a means of high-resolution, low-damage imaging of radiation-sensitive samples, such as biomolecules and proteins. The basic setup for IFM is a Mach-Zehnder interferometer, and recent progress in nanofabricated electron diffraction gratings has made it possible to incorporate a Mach-Zehnder interferometer in a transmission-electron microscope (TEM). Therefore, the limits of performance of IFM with such an interferometer and a shot-noise limited electron source (such as that in a TEM) are of interest. In this work, we compared the error probability and sample damage for ideal IFM and classical imaging schemes, through theoretical analysis and numerical simulation. We considered a sample that is either completely transparent or completely opaque at each pixel. In our analysis, we also evaluated the impact of an additional detector for scattered electrons. The additional detector resulted in reduction of error by up to an order of magnitude, for both IFM and classical schemes. We also investigated a sample re-illumination scheme based on updating priors after each round of illumination and found that this scheme further reduced error by a factor of two. Implementation of these methods is likely achievable with existing instrumentation and would result in improved resolution in low-dose electron microscopy.

physics.ins-det↗

arXiv2019

A General Theoretical and Experimental Framework for Nanoscale Electromagnetism

Local, bulk response functions, e.g permittivity, and the macroscopic Maxwell equations completely specify the classical electromagnetic problem, which features only wavelength $λ$ and geometric scales. The above neglect of intrinsic electronic length scales $L_{\text{e}}$ leads to an eventual breakdown in the nanoscopic limit. Here, we present a general theoretical and experimental framework for treating nanoscale electromagnetic phenomena. The framework features surface-response functions---known as the Feibelman $d$-parameters---which reintroduce the missing electronic length scales. As a part of our framework, we establish an experimental procedure to measure these complex, dispersive surface response functions, enabled by quasi-normal-mode perturbation theory and observations of pronounced nonclassical effects---spectral shifts in excess of 30% and the breakdown of Kreibig-like broadening---in a quintessential multiscale architecture: film-coupled nanoresonators, with feature-sizes comparable to both $L_{\text{e}}$ and $λ$.

physics.optics↗

arXiv2018

Enhancement of optical response in nanowires by negative-tone PMMA lithography

The method of negative-tone-PMMA electron-beam lithography is investigated to improve the performance of nanowire-based superconducting detectors. Using this approach, the superconducting nanowire single-photon detectors (SNSPDs) have been fabricated from thick 5-nm NbN film sputtered at the room temperature. To investigate the impact of this process, SNSPDs were prepared by positive-tone and negative-tone-PMMA lithography, and their electrical and photodetection characteristics at 4.2 K were compared. The SNSPDs made by negative-tone-PMMA lithography show higher critical-current density and higher photon count rate at various wavelengths. Our results suggest a higher negative-tone-PMMA technology may be preferable to the standard positive-tone-PMMA lithography for this application.

physics.app-ph↗

arXiv2018

Unravelling Robustness of Deep Learning based Face Recognition Against Adversarial Attacks

Deep neural network (DNN) architecture based models have high expressive power and learning capacity. However, they are essentially a black box method since it is not easy to mathematically formulate the functions that are learned within its many layers of representation. Realizing this, many researchers have started to design methods to exploit the drawbacks of deep learning based algorithms questioning their robustness and exposing their singularities. In this paper, we attempt to unravel three aspects related to the robustness of DNNs for face recognition: (i) assessing the impact of deep architectures for face recognition in terms of vulnerabilities to attacks inspired by commonly observed distortions in the real world that are well handled by shallow learning methods along with learning based adversaries; (ii) detecting the singularities by characterizing abnormal filter response behavior in the hidden layers of deep networks; and (iii) making corrections to the processing pipeline to alleviate the problem. Our experimental evaluation using multiple open-source DNN-based face recognition networks, including OpenFace and VGG-Face, and two publicly available databases (MEDS and PaSC) demonstrates that the performance of deep learning based face recognition algorithms can suffer greatly in the presence of such distortions. The proposed method is also compared with existing detection algorithms and the results show that it is able to detect the attacks with very high accuracy by suitably designing a classifier using the response of the hidden layers in the network. Finally, we present several effective countermeasures to mitigate the impact of adversarial attacks and improve the overall robustness of DNN-based face recognition.

cs.CV↗

arXiv2016

A nanofabricated, monolithic, path-separated electron interferometer

We report a self-aligned, monolithic electron interferometer, consisting of two 45 nm thick silicon layers separated by 20 $μ$m. This interferometer was fabricated from a single crystal silicon cantilever on a transmission electron microscope grid by gallium focused ion-beam milling. Using this interferometer, we demonstrate beam path-separation, and obtain interference fringes in a Mach-Zehnder geometry, in an unmodified 200 kV transmission electron microscope. The fringes have a period of 0.32 nm, which corresponds to the $\left[\bar{1}\bar{1}1\right]$ lattice planes of silicon, and a maximum contrast of 15 %. This design can potentially be scaled to millimeter-scale, and used in electron holography. It can also be applied to perform fundamental physics experiments, such as interaction-free measurement with electrons.

physics.ins-det↗

arXiv2014

Power and efficiency analysis of a realistic resonant tunneling diode thermoelectric

Low-dimensional systems with sharp features in the density of states have been proposed as a means to improving the efficiency of thermoelectric devices. Quantum dot systems, which offer the sharpest density of states achievable, however, suffer from low power outputs while bulk (3-D) thermoelectrics, while displaying high power outputs, offer very low efficiencies. Here, we analyze the use of a resonant tunneling diode structure that combines the best of both aspects, that is, density of states distortion with a finite bandwidth due to confinement that aids the efficiency and a large number of current carrying transverse modes that enhances the total power output. We show that this device can achieve a high power output ($\sim 0.3$ MW$/m^2$) at efficiencies of $\sim 40\%$ of the Carnot efficiency due to the contribution from these transverse momentum states at a finite bandwidth of $kT/2$. We then provide a detailed analysis of the physics of charge and heat transport with insights on parasitic currents that reduce the efficiency. Finally, a comparison between the resonant tunneling diode and a quantum dot device with comparable bandwidth reveals that a similar performance requires ultra-dense areal quantum dot densities of $\sim 10^{12}/cm^2$.

cond-mat.mes-hall↗