arXiv subjects

Abhishek Kumar

Publications and source records attributed to Abhishek Kumar.

At least 145 records · Page 8Linked to original sources

arXiv2021

Retrieve in Style: Unsupervised Facial Feature Transfer and Retrieval

We present Retrieve in Style (RIS), an unsupervised framework for facial feature transfer and retrieval on real images. Recent work shows capabilities of transferring local facial features by capitalizing on the disentanglement property of the StyleGAN latent space. RIS improves existing art on the following: 1) Introducing more effective feature disentanglement to allow for challenging transfers (ie, hair, pose) that were not shown possible in SoTA methods. 2) Eliminating the need for per-image hyperparameter tuning, and for computing a catalog over a large batch of images. 3) Enabling fine-grained face retrieval using disentangled facial features (eg, eyes). To our best knowledge, this is the first work to retrieve face images at this fine level. 4) Demonstrating robust, natural editing on real images. Our qualitative and quantitative analyses show RIS achieves both high-fidelity feature transfers and accurate fine-grained retrievals on real images. We also discuss the responsible applications of RIS.

cs.CV↗

arXiv2021

Zero-field spin resonance in graphene with proximity-induced spin-orbit coupling

We investigate collective spin excitations in graphene with proximity-induced spin-orbit coupling (SOC) of the Rashba and valley-Zeeman types, as it is the case, e.g., for graphene on transition- metal-dichalcogenide substrates. It is shown that, even in the absence of an external magnetic field, such a system supports collective modes, which correspond to coupled oscillations of the uniform and valley-staggered magnetizations. These modes can be detected via both zero-field electron spin resonance (ESR) and zero-field electric-dipole spin resonance (EDSR), with EDSR response coming solely from Rashba SOC. We analyze the effect of electron-electron interaction within the Fermi- liquid kinetic equation and show that the interaction splits both the ESR and EDSR peaks into two. The magnitude of splitting and the relative weights of the resonances can be used to extract the spin-orbit coupling constants and many-body interaction parameters that may not be accessible by other methods.

cond-mat.str-el↗

arXiv2021

Implicit Rate-Constrained Optimization of Non-decomposable Objectives

We consider a popular family of constrained optimization problems arising in machine learning that involve optimizing a non-decomposable evaluation metric with a certain thresholded form, while constraining another metric of interest. Examples of such problems include optimizing the false negative rate at a fixed false positive rate, optimizing precision at a fixed recall, optimizing the area under the precision-recall or ROC curves, etc. Our key idea is to formulate a rate-constrained optimization that expresses the threshold parameter as a function of the model parameters via the Implicit Function theorem. We show how the resulting optimization problem can be solved using standard gradient based methods. Experiments on benchmark datasets demonstrate the effectiveness of our proposed method over existing state-of-the art approaches for these problems. The code for the proposed method is available at https://github.com/google-research/google-research/tree/master/implicit_constrained_optimization .

cs.LG↗

arXiv2021

Wavelet Design in a Learning Framework

Wavelets have proven to be highly successful in several signal and image processing applications. Wavelet design has been an active field of research for over two decades, with the problem often being approached from an analytical perspective. In this paper, we introduce a learning based approach to wavelet design. We draw a parallel between convolutional autoencoders and wavelet multiresolution approximation, and show how the learning angle provides a coherent computational framework for addressing the design problem. We aim at designing data-independent wavelets by training filterbank autoencoders, which precludes the need for customized datasets. In fact, we use high-dimensional Gaussian vectors for training filterbank autoencoders, and show that a near-zero training loss implies that the learnt filters satisfy the perfect reconstruction property with very high probability. Properties of a wavelet such as orthogonality, compact support, smoothness, symmetry, and vanishing moments can be incorporated by designing the autoencoder architecture appropriately and with a suitable regularization term added to the mean-squared error cost used in the learning process. Our approach not only recovers the well known Daubechies family of orthogonal wavelets and the Cohen-Daubechies-Feauveau family of symmetric biorthogonal wavelets, but also learns wavelets outside these families.

cs.LG↗

arXiv2021

Risk score learning for COVID-19 contact tracing apps

Digital contact tracing apps for COVID, such as the one developed by Google and Apple, need to estimate the risk that a user was infected during a particular exposure, in order to decide whether to notify the user to take precautions, such as entering into quarantine, or requesting a test. Such risk score models contain numerous parameters that must be set by the public health authority. In this paper, we show how to automatically learn these parameters from data. Our method needs access to exposure and outcome data. Although this data is already being collected (in an aggregated, privacy-preserving way) by several health authorities, in this paper we limit ourselves to simulated data, so that we can systematically study the different factors that affect the feasibility of the approach. In particular, we show that the parameters become harder to estimate when there is more missing data (e.g., due to infections which were not recorded by the app), and when there is model misspecification. Nevertheless, the learning approach outperforms a strong manually designed baseline. Furthermore, the learning approach can adapt even when the risk factors of the disease change, e.g., due to the evolution of new variants, or the adoption of vaccines.

cs.LG↗

arXiv2021

SOCluster- Towards Intent-based Clustering of Stack Overflow Questions using Graph-Based Approach

Stack Overflow (SO) platform has a huge dataset of questions and answers driven by interactions between users. But the count of unanswered questions is continuously rising. This issue is common across various community Question & Answering platforms (Q&A) such as Yahoo, Quora and so on. Clustering is one of the approaches used by these communities to address this challenge. Specifically, Intent-based clustering could be leveraged to answer unanswered questions using other answered questions in the same cluster and can also improve the response time for new questions. It is here, we propose SOCluster, an approach and a tool to cluster SO questions based on intent using a graph-based clustering approach. We selected four datasets of 10k, 20k, 30k & 40k SO questions without code-snippets or images involved, and performed intent-based clustering on them. We have done a preliminary evaluation of our tool by analyzing the resultant clusters using the commonly used metrics of Silhouette coefficient, Calinkski-Harabasz Index, & Davies-Bouldin Index. We performed clustering for 8 different threshold similarity values and analyzed the intriguing trends reflected by the output clusters through the three evaluation metrics. At 90% threshold similarity, it shows the best value for the three evaluation metrics on all four datasets. The source code and tool are available for download on Github at: https://github.com/Liveitabhi/SOCluster, and the demo can be found here: https://youtu.be/uyn8ie4h3NY.

cs.SE↗

arXiv2021

Floquet Gauge Pumps as Sensors for Spectral Degeneracies Protected by Symmetry or Topology

We introduce the concept of a Floquet gauge pump whereby a dynamically engineered Floquet Hamiltonian is employed to reveal the inherent degeneracy of the ground state in interacting systems. We demonstrate this concept in a one-dimensional XY model with periodically driven couplings and transverse field. In the high-frequency limit, we obtain the Floquet Hamiltonian consisting of the static XY and dynamically generated Dzyaloshinsky-Moriya interaction (DMI) terms. The dynamically generated magnetization current depends on the phases of complex coupling terms, with the XY interaction as the real and DMI as the imaginary part. As these phases are cycled, the current reveals the ground-state degeneracies that distinguish the ordered and disordered phases. We discuss experimental requirements needed to realize the Floquet gauge pump in a synthetic quantum spin system of interacting trapped ions.

cond-mat.quant-gas↗

arXiv2021

Quasiparticle and Nonquasiparticle Transport in Doped Quantum Paraelectrics

Charge transport in doped quantum paralectrics (QPs) presents a number of puzzles, including a pronounced $T^2$ regime in the resistivity. We analyze charge transport in a QP within a model of electrons coupled to a soft transverse optical (TO) mode via a two-phonon mechanism. For $T$ above the soft-mode frequency but below some characteristic scale ($E_0$), the resistivity scales with the occupation number of phonons squared, i.e., as $T^2$. The $T^2$ scattering rate does not depend on the carrier number density and is not affected by a crossover between degenerate and non-degenerate regimes, in agreement with the experiment. Temperatures higher than $E_0$ correspond to a non-quasiparticle regime, which we analyze by mapping the Dyson equation onto a problem of supersymmetric quantum mechanics. The combination of scattering by two TO phonons and by a longitudinal optical mode explains the data quite well.

cond-mat.str-el↗

arXiv2021

Score-Based Generative Modeling through Stochastic Differential Equations

Creating noise from data is easy; creating data from noise is generative modeling. We present a stochastic differential equation (SDE) that smoothly transforms a complex data distribution to a known prior distribution by slowly injecting noise, and a corresponding reverse-time SDE that transforms the prior distribution back into the data distribution by slowly removing the noise. Crucially, the reverse-time SDE depends only on the time-dependent gradient field (\aka, score) of the perturbed data distribution. By leveraging advances in score-based generative modeling, we can accurately estimate these scores with neural networks, and use numerical SDE solvers to generate samples. We show that this framework encapsulates previous approaches in score-based generative modeling and diffusion probabilistic modeling, allowing for new sampling procedures and new modeling capabilities. In particular, we introduce a predictor-corrector framework to correct errors in the evolution of the discretized reverse-time SDE. We also derive an equivalent neural ODE that samples from the same distribution as the SDE, but additionally enables exact likelihood computation, and improved sampling efficiency. In addition, we provide a new way to solve inverse problems with score-based models, as demonstrated with experiments on class-conditional generation, image inpainting, and colorization. Combined with multiple architectural improvements, we achieve record-breaking performance for unconditional image generation on CIFAR-10 with an Inception score of 9.89 and FID of 2.20, a competitive likelihood of 2.99 bits/dim, and demonstrate high fidelity generation of 1024 x 1024 images for the first time from a score-based generative model.

cs.LG↗

arXiv2020

On Implicit Regularization in $β$-VAEs

While the impact of variational inference (VI) on posterior inference in a fixed generative model is well-characterized, its role in regularizing a learned generative model when used in variational autoencoders (VAEs) is poorly understood. We study the regularizing effects of variational distributions on learning in generative models from two perspectives. First, we analyze the role that the choice of variational family plays in imparting uniqueness to the learned model by restricting the set of optimal generative models. Second, we study the regularization effect of the variational family on the local geometry of the decoding model. This analysis uncovers the regularizer implicit in the $β$-VAE objective, and leads to an approximation consisting of a deterministic autoencoding objective plus analytic regularizers that depend on the Hessian or Jacobian of the decoding model, unifying VAEs with recent heuristics proposed for training regularized autoencoders. We empirically verify these findings, observing that the proposed deterministic objective exhibits similar behavior to the $β$-VAE in terms of objective value and sample quality.

cs.LG↗

arXiv2020

Generalized Adversarially Learned Inference

Allowing effective inference of latent vectors while training GANs can greatly increase their applicability in various downstream tasks. Recent approaches, such as ALI and BiGAN frameworks, develop methods of inference of latent variables in GANs by adversarially training an image generator along with an encoder to match two joint distributions of image and latent vector pairs. We generalize these approaches to incorporate multiple layers of feedback on reconstructions, self-supervision, and other forms of supervision based on prior or learned knowledge about the desired solutions. We achieve this by modifying the discriminator's objective to correctly identify more than two joint distributions of tuples of an arbitrary number of random variables consisting of images, latent vectors, and other variables generated through auxiliary tasks, such as reconstruction and inpainting or as outputs of suitable pre-trained models. We design a non-saturating maximization objective for the generator-encoder pair and prove that the resulting adversarial game corresponds to a global optimum that simultaneously matches all the distributions. Within our proposed framework, we introduce a novel set of techniques for providing self-supervised feedback to the model based on properties, such as patch-level correspondence and cycle consistency of reconstructions. Through comprehensive experiments, we demonstrate the efficacy, scalability, and flexibility of the proposed approach for a variety of tasks.

cs.LG↗

arXiv2020

A Review of the Anomalies in Directed Energy Deposition (DED) Processes and Potential Solutions

Directed Energy Deposition (DED) processes offer one of the most versatile current techniques to additively manufacture and repair metallic components that have generatively designed complex geometries, and with compositional control. When compared to powder bed fusion (PBF), its applicability and adoption has been limited because several issues innate to the process are yet to be suitably understood and resolved. This work catalogs and delineates these issues and anomalies in the DED process along with their causes and solutions, based on a state-of-the-art literature review. This work also serves to enumerate and associate the underlying causes to the detrimental effects which manifest as undesirable part/process outcomes. These DED-specific anomalies are categorized under groups related to the part, process, material, productivity, safety, repair, and functional gradients. Altogether, this primer acts as a guide to best prepare for and mitigate the problems that are encountered in DED, and also to lay the groundwork to inspire novel solutions to further advance DED into mainstream manufacturing.

physics.app-ph↗

arXiv2020

Adaptive Contention Window Design using Deep Q-learning

We study the problem of adaptive contention window (CW) design for random-access wireless networks. More precisely, our goal is to design an intelligent node that can dynamically adapt its minimum CW (MCW) parameter to maximize a network-level utility knowing neither the MCWs of other nodes nor how these change over time. To achieve this goal, we adopt a reinforcement learning (RL) framework where we circumvent the lack of system knowledge with local channel observations and we reward actions that lead to high utilities. To efficiently learn these preferred actions, we follow a deep Q-learning approach, where the Q-value function is parametrized using a multi-layer perception. In particular, we implement a rainbow agent, which incorporates several empirical improvements over the basic deep Q-network. Numerical experiments based on the NS3 simulator reveal that the proposed RL agent performs close to optimal and markedly improves upon existing learning and non-learning based alternatives.

eess.SP↗

arXiv2020

Few-Shot Adaptation of Generative Adversarial Networks

Generative Adversarial Networks (GANs) have shown remarkable performance in image synthesis tasks, but typically require a large number of training samples to achieve high-quality synthesis. This paper proposes a simple and effective method, Few-Shot GAN (FSGAN), for adapting GANs in few-shot settings (less than 100 images). FSGAN repurposes component analysis techniques and learns to adapt the singular values of the pre-trained weights while freezing the corresponding singular vectors. This provides a highly expressive parameter space for adaptation while constraining changes to the pretrained weights. We validate our method in a challenging few-shot setting of 5-100 images in the target domain. We show that our method has significant visual quality gains compared with existing GAN adaptation methods. We report qualitative and quantitative results showing the effectiveness of our method. We additionally highlight a problem for few-shot synthesis in the standard quantitative metric used by data-efficient image synthesis works. Code and additional results are available at http://e-271.github.io/few-shot-gan.

cs.CV↗

arXiv2020

Temperature Dependent Non-linear Damping in Palladium Nano-mechanical Resonators

Advances in nano-fabrication techniques has made it feasible to observe damping phenomena beyond the linear regime in nano-mechanical systems. In this work, we report cubic non-linear damping in palladium nano-mechanical resonators. Nano-scale palladium beams exposed to a $H_2$ atmosphere become softer and display enhanced Duffing non-linearity as well as non-linear damping at ultra low temperatures. The damping is highest at the lowest temperatures of $\sim 110\: mK$ and decreases when warmed up-to $\sim 1\textrm{ }K$. We experimentally demonstrate for the first time a temperature dependent non-linear damping in a nano-mechanical system below 1 K. It is consistent with a predicted two phonon mediated non-linear Akhiezer scenario for ballistic phonons with mean free path comparable to the beam thickness. This opens up new possibilities to engineer non-linear phenomena at low temperatures.

cond-mat.mes-hall↗

arXiv2020

Microstructure evolution and densification during spark plasma sintering of nanocrystalline W-5wt.%Ta alloy

The present work reports the effect of Ta on densification and microstructure evolution during non-isothermal and spark plasma sintering of nanocrystalline W. Nanocrystalline W-5wt.%Ta alloy powder was synthesized using mechanical alloying. The nanocrystalline powder was characterized thoroughly using X-ray diffraction line profile analysis. Furthermore, the shrinkage behavior of nanocrystalline powder was investigated during non-isothermal sintering using dilatometry. Subsequently, the alloy powder was consolidated using spark plasma sintering up to 1600 °C. The role of Ta on stabilizing the microstructure during spark plasma sintering of nanocrystalline W was investigated in detail using electron backscatter diffraction. The average grain size of spark plasma sintered W-5wt.%Ta alloy was observed as 1.73 micron.

cond-mat.mtrl-sci↗

arXiv2020

Linear Response Theory and Optical Conductivity of Floquet Topological Insulators

Motivated by the quest for experimentally accessible dynamical probes of Floquet topological insulators, we formulate the linear response theory of a periodically driven system. We illustrate the applications of this formalism by giving general expressions for optical conductivity of Floquet systems, including its homodyne and heterodyne components and beyond. We obtain the Floquet optical conductivity of specific driven models, including two-dimensional Dirac material such as the surface of a topological insulator, graphene, and the Haldane model irradiated with circularly or linearly polarized laser, as well as semiconductor quantum well driven by an ac potential. We obtain approximate analytical expressions and perform numerically exact calculations of the Floquet optical conductivity in different scenarios of the occupation of the Floquet bands, in particular, the diagonal Floquet distribution and the distribution obtained after a quench. We comment on experimental signatures and detection of Floquet topological phases using optical probes.

cond-mat.mes-hall↗

arXiv2020

MCQA: Multimodal Co-attention Based Network for Question Answering

We present MCQA, a learning-based algorithm for multimodal question answering. MCQA explicitly fuses and aligns the multimodal input (i.e. text, audio, and video), which forms the context for the query (question and answer). Our approach fuses and aligns the question and the answer within this context. Moreover, we use the notion of co-attention to perform cross-modal alignment and multimodal context-query alignment. Our context-query alignment module matches the relevant parts of the multimodal context and the query with each other and aligns them to improve the overall performance. We evaluate the performance of MCQA on Social-IQ, a benchmark dataset for multimodal question answering. We compare the performance of our algorithm with prior methods and observe an accuracy improvement of 4-7%.

cs.CL↗