Searcharxiv⌕ Search

arXiv subjects

Abhishek Sharma

Publications and source records attributed to Abhishek Sharma.

At least 73 records · Page 4Linked to original sources
arXiv2024

Autonomous programmable microscopic electronic lablets optimized with digital control

Lablets are autonomous microscopic particles with programmable CMOS electronics that can control electrokinetic phenomena and electrochemical reactions in solution via actuator and sensor microelectrodes. In this paper, we describe the design and fabrication of optimized singulated lablets (CMOS3) with dimensions 140x140x50 micrometers carrying an integrated coplanar encapsulated supercapacitor as a rechargeable power supply. The lablets are designed to allow docking to one another or to a smart surface for interchange of energy, electronic information, and chemicals. The paper focusses on the digital and analog design of the lablets to allow significant programmable functionality in a microscopic footprint, including the control of autonomous actuation and sensing up to the level of being able to support a complete lablet self-reproduction life cycle, although experimentally this remains to be proven. The potential of lablets in autonomous sensing and control and for evolutionary experimentation are discussed.

cs.RO↗
arXiv2024

Design and fabrication of autonomous electronic lablets for chemical control

Lablets are autonomous microscopic particles with programmable CMOS electronics that canvcontrol electrokinetic phenomena and electrochemical reactions in solution via actuator and sensor microelectrodes. The lablets are designed to be rechargeable using an integrated supercapacitor, and to allow docking to one another or to a smart surface for interchange of energy, electronic information and chemicals. In this paper, we describe the design and fabrication of singulated lablets (CMOS2) at the scale of 100 by 200 μm, with the supercap adjacent to the functional lablet and occupying half the space. In other works, we have characterized the supercap and described the electronic design and proven functionality using arrays of these lablets. Here we present fabrication details for integrating functional coatings and the supercap and demonstrate electronic functionality of the lablets following singulation.

cs.RO↗
arXiv2024

Quantum-Accurate Machine Learning Potentials for Metal-Organic Frameworks using Temperature Driven Active Learning

Understanding how structural flexibility affects the properties of metal-organic frameworks (MOFs) is crucial for the design of better MOFs for targeted applications. Flexible MOFs can be studied with molecular dynamics simulations, whose accuracy depends on the force-field used to describe the interatomic interactions. Density functional theory (DFT) and quantum-chemistry methods are highly accurate, but the computational overheads limit their use in long time-dependent simulations for large systems. In contrast, classical force fields usually struggle with the description of coordination bonds. In this work we develop a DFT-accurate machine-learning spectral neighbor analysis potential, trained on DFT energies, forces and stress tensors, for two representative MOFs, namely ZIF-8 and MOF-5. Their structural and vibrational properties are then studied as a function of temperature and tightly compared with available experimental data. Most importantly, we demonstrate an active-learning algorithm, based on mapping the relevant internal coordinates, which drastically reduces the number of training data to be computed at the DFT level. Thus, the workflow presented here appears as an efficient strategy for the study of flexible MOFs with DFT accuracy, but at a fraction of the DFT computational cost.

cond-mat.mtrl-sci↗
arXiv2024

Re-DiffiNet: Modeling discrepancies in tumor segmentation using diffusion models

Identification of tumor margins is essential for surgical decision-making for glioblastoma patients and provides reliable assistance for neurosurgeons. Despite improvements in deep learning architectures for tumor segmentation over the years, creating a fully autonomous system suitable for clinical floors remains a formidable challenge because the model predictions have not yet reached the desired level of accuracy and generalizability for clinical applications. Generative modeling techniques have seen significant improvements in recent times. Specifically, Generative Adversarial Networks (GANs) and Denoising-diffusion-based models (DDPMs) have been used to generate higher-quality images with fewer artifacts and finer attributes. In this work, we introduce a framework called Re-Diffinet for modeling the discrepancy between the outputs of a segmentation model like U-Net and the ground truth, using DDPMs. By explicitly modeling the discrepancy, the results show an average improvement of 0.55\% in the Dice score and 16.28\% in HD95 from cross-validation over 5-folds, compared to the state-of-the-art U-Net segmentation model.

eess.IV↗
arXiv2024

Ultrahigh Frequency and Multi-channel Output in Skyrmion Based Nano-oscillator

Spintronic nano-oscillators can generate tunable microwave signals that find a wide range of applications in the field of telecommunication to modern neuromorphic computing systems. Among other spintronic devices, a magnetic skyrmion is a promising candidate for the next generation of low-power devices due to its small size and topological stability. In this work, we propose a multi-channel oscillator design based on the synthetic anti-ferromagnetic (SAF) skyrmion pair. The mitigation of the skyrmion Hall effect in SAF and the associated decimation of the Magnus force endows the proposed oscillator with an ultra-high frequency of 41GHz and a multi-channel frequency output driven by the same current. The ultrahigh operational frequency represents an $\sim$342 times improvement compared to the monolayer single skyrmion oscillator featuring a constant uniaxial anisotropy profile. Using micromagnetic simulations, we demonstrate the effectiveness of our proposed multi-channel oscillator design by introducing multi-channel nanotracks along with multiple skyrmions for enhanced frequency operation. The ultrahigh operational frequency and multi-channel output are attributed to three key factors: The oscillator design accounting for a finite spin-flip length of the spacer (such as Ru) material, tangential velocity proportionality on input spin current along with weak dependence on the radius of rotation of the skyrmion-pair, skyrmion interlocking in the channel enabled by the multi-channel high Ku rings and skyrmion-skyrmion repulsion, therefore resulting ultrahigh frequency and multi-channel outputs.

cond-mat.mes-hall↗
arXiv2024

PBADet: A One-Stage Anchor-Free Approach for Part-Body Association

The detection of human parts (e.g., hands, face) and their correct association with individuals is an essential task, e.g., for ubiquitous human-machine interfaces and action recognition. Traditional methods often employ multi-stage processes, rely on cumbersome anchor-based systems, or do not scale well to larger part sets. This paper presents PBADet, a novel one-stage, anchor-free approach for part-body association detection. Building upon the anchor-free object representation across multi-scale feature maps, we introduce a singular part-to-body center offset that effectively encapsulates the relationship between parts and their parent bodies. Our design is inherently versatile and capable of managing multiple parts-to-body associations without compromising on detection accuracy or robustness. Comprehensive experiments on various datasets underscore the efficacy of our approach, which not only outperforms existing state-of-the-art techniques but also offers a more streamlined and efficient solution to the part-body association challenge.

cs.CV↗
arXiv2024

An Optimization Framework for Processing and Transfer Learning for the Brain Tumor Segmentation

Tumor segmentation from multi-modal brain MRI images is a challenging task due to the limited samples, high variance in shapes and uneven distribution of tumor morphology. The performance of automated medical image segmentation has been significant improvement by the recent advances in deep learning. However, the model predictions have not yet reached the desired level for clinical use in terms of accuracy and generalizability. In order to address the distinct problems presented in Challenges 1, 2, and 3 of BraTS 2023, we have constructed an optimization framework based on a 3D U-Net model for brain tumor segmentation. This framework incorporates a range of techniques, including various pre-processing and post-processing techniques, and transfer learning. On the validation datasets, this multi-modality brain tumor segmentation framework achieves an average lesion-wise Dice score of 0.79, 0.72, 0.74 on Challenges 1, 2, 3 respectively.

eess.IV↗
arXiv2023

Numerical Investigation of Supercritical Combustion Dynamics in a Multi-Element LOx-Methane Combustor Using Flamelet-Generated Manifold Approach

The article investigates liquid oxygen (LOx)-methane supercritical combustion dynamics in a multi-element rocket scale combustor using large eddy simulation (LES). A complex framework of real gas thermodynamics and flamelet generated manifold (FGM) combustion model is invoked to simulate transcritical oxygen injection and supercritical methane combustion. A benchmark Mascotte chamber, Rocket Combustion Modelling (RCM) Test Case, i.e. RCM-3 (V04)/G2 test case, is used to validate the real gas FGM model in the LES framework. The validation study accurately reproduces experimental flame structure and OH concentration, demonstrating the FGM model's importance in incorporating finite rate kinetics in LOx-methane combustion. Subsequently, the numerical framework investigates a specially designed multi-element combustion chamber featuring seven bidirectional swirl coaxial injectors. The analyses capture the complex hydrodynamics and combustion dynamics associated with multiple swirl injectors operating at supercritical pressure, effectively demonstrating the initiation of transverse acoustic waves and examining the effect of local sound speed on the evolution of acoustic modes in the combustor. The dominant frequency modes shed light on understanding the role of injectors in enhanced combustor dynamics. Spectral analysis reveals the interplay of the upstream injector and chamber acoustics due to possible frequency coupling. The results also highlight the effect of fuel injection temperature on the stability of the combustor, revealing a violent dynamic activity for lower fuel injection temperature associated with the longitudinal acoustic mode of the combustor. The investigation appropriately reproduces self-sustained limit cycle oscillations at lower fuel injection temperatures and corroborates the conventional understanding of combustor instability.

physics.flu-dyn↗
arXiv2023

Determining Molecular Complexity using Assembly Theory and Spectroscopy

Determining the complexity of molecules has important applications from molecular design to understanding the history of the process that led to the formation of the molecule. Currently, it is not possible to experimentally determine, without full structure elucidation, how complex a molecule is. Assembly Theory has been developed to quantify the complexity of a molecule by finding the shortest path to construct the molecule from building blocks, revealing its molecular assembly index (MA). In this study, we present an approach to rapidly and exhaustively calculate the MA of molecules from the spectroscopic measurements. We demonstrate that molecular complexity (MA) can be experimentally estimated using three independent techniques: nuclear magnetic resonance (NMR), tandem mass spectrometry (MS/MS), and infrared spectroscopy (IR), and these give consistent results with good correlations with the theoretically determined values from assembly theory. By identifying and analysing the number of absorbances in IR spectra, carbon resonances in NMR, or molecular fragments in tandem MS, the molecular assembly of an unknown molecule can be reliably estimated from experimental data. This represents the first experimentally quantifiable approach to defining molecular assembly, a reliable metric for complexity, as an intrinsic property of molecules and can also be performed on complex mixtures. This paves the way to use spectroscopic and spectrometric techniques to unambiguously detect alien life in the solar system, and beyond on exoplanets.

q-bio.QM↗
arXiv2023

Reinforced Self-Training (ReST) for Language Modeling

Reinforcement learning from human feedback (RLHF) can improve the quality of large language model's (LLM) outputs by aligning them with human preferences. We propose a simple algorithm for aligning LLMs with human preferences inspired by growing batch reinforcement learning (RL), which we call Reinforced Self-Training (ReST). Given an initial LLM policy, ReST produces a dataset by generating samples from the policy, which are then used to improve the LLM policy using offline RL algorithms. ReST is more efficient than typical online RLHF methods because the training dataset is produced offline, which allows data reuse. While ReST is a general approach applicable to all generative learning settings, we focus on its application to machine translation. Our results show that ReST can substantially improve translation quality, as measured by automated metrics and human evaluation on machine translation benchmarks in a compute and sample-efficient manner.

cs.CL↗
arXiv2023

Performance of the MALTA Telescope

MALTA is part of the Depleted Monolithic Active Pixel sensors designed in Tower 180nm CMOS imaging technology. A custom telescope with six MALTA planes has been developed for test beam campaigns at SPS, CERN, with the ability to host several devices under test. The telescope system has a dedicated custom readout, online monitoring integrated into DAQ with realtime hit map, time distribution and event hit multiplicity. It hosts a dedicated fully configurable trigger system enabling to trigger on coincidence between telescope planes and timing reference from a scintillator. The excellent time resolution performance allows for fast track reconstruction, due to the possibility to retain a low hit multiplicity per event which reduces the combinatorics. This paper reviews the architecture of the system and its performance during the 2021 and 2022 test beam campaign at the SPS North Area.

hep-ex↗
arXiv2023

Performance Study of Strongly Coupled Magnetic Resonance

Strongly Coupled Magnetic Resonance (SCMR) uses electromagnetic resonance in order to efficiently transfer power wirelessly over mid-range distances. Since the energy exchange capability of resonant objects higher than non-resonant objects, strongly coupled systems are able to achieve more efficient energy transfer than other wireless power transfer systems. The paper presents detailed experimental and simulated analysis of the performance of the SCMR system. A prototype of the SCMR system was implemented and experiments were conducted to analyze the performance of the system. Finally, the resonant frequency of the system was experimentally verified and the factors influencing the wireless power transfer were also studied

physics.app-ph↗
arXiv2023

Object-Aware Cropping for Self-Supervised Learning

A core component of the recent success of self-supervised learning is cropping data augmentation, which selects sub-regions of an image to be used as positive views in the self-supervised loss. The underlying assumption is that randomly cropped and resized regions of a given image share information about the objects of interest, which the learned representation will capture. This assumption is mostly satisfied in datasets such as ImageNet where there is a large, centered object, which is highly likely to be present in random crops of the full image. However, in other datasets such as OpenImages or COCO, which are more representative of real world uncurated data, there are typically multiple small objects in an image. In this work, we show that self-supervised learning based on the usual random cropping performs poorly on such datasets. We propose replacing one or both of the random crops with crops obtained from an object proposal algorithm. This encourages the model to learn both object and scene level semantic representations. Using this approach, which we call object-aware cropping, results in significant improvements over scene cropping on classification and object detection benchmarks. For example, on OpenImages, our approach achieves an improvement of 8.8% mAP over random scene-level cropping using MoCo-v2 based pre-training. We also show significant improvements on COCO and PASCAL-VOC object detection and segmentation tasks over the state-of-the-art self-supervised learning approaches. Our approach is efficient, simple and general, and can be used in most existing contrastive and non-contrastive self-supervised learning frameworks.

cs.CV↗
arXiv2023

Non-uniform Superlattice Magnetic Tunnel Junctions

We propose a new class of non-uniform superlattice magnetic tunnel junctions (Nu-SLTJs) with the Linear, Gaussian, Lorentzian, and Pöschl-teller width and height based profiles manifesting a sizable enhancement in the TMR($\approx 10^4-10^6\%$) with a significant suppression in the switching bias($\approx$9 folds) owing to the physics of broad-band spin filtering. By exploring the negative differential resistance region in the current-voltage characteristics of the various Nu-SLTJs, we predict the Nu-SLTJs offer the fastest spin transfer torque switching in the order of a few hundred picoseconds. We self-consistently employ the atomistic non-equilibrium Green's function formalism coupled with the Landau-Lifshitz-Gilbert-Slonczewski equation to evaluate the device performance of the various Nu-SLTJs. We also present the design of minimal three-barrier Nu-SLTJs having significant TMR($\approx 10^4\%$) and large spin current for ease of device fabrication. We hope that the class of Nu-SLTJs proposed in this work may lay the bedrock to embark on the exhilarating voyage of exploring various non-uniform superlattices for the next generation of spintronic devices.

cond-mat.mes-hall↗
arXiv2023

Assembly Theory Explains and Quantifies the Emergence of Selection and Evolution

Since the time of Darwin, scientists have struggled to reconcile the evolution of biological forms in a universe determined by fixed laws. These laws underpin the origin of life, evolution, human culture and technology, as set by the boundary conditions of the universe, however these laws cannot predict the emergence of these things. By contrast evolutionary theory works in the opposite direction, indicating how selection can explain why some things exist and not others. To understand how open-ended forms can emerge in a forward-process from physics that does not include their design, a new approach to understand the non-biological to biological transition is necessary. Herein, we present a new theory, Assembly Theory (AT), which explains and quantifies the emergence of selection and evolution. In AT, the complexity of an individual observable object is measured by its Assembly Index (a), defined as the minimal number of steps needed to construct the object from basic building blocks. Combining a with the copy number defines a new quantity called Assembly which quantifies the amount of selection required to produce a given ensemble of objects. We investigate the internal structure and properties of assembly space and quantify the dynamics of undirected exploratory processes as compared to the directed processes that emerge from selection. The implementation of assembly theory allows the emergence of selection in physical systems to be quantified at any scale as the transition from undirected-discovery dynamics to a selected process within the assembly space. This yields a mechanism for the onset of selection and evolution and a formal approach to defining life. Because the assembly of an object is easily calculable and measurable it is possible to quantify a lower limit on the amount of selection and memory required to produce complexity uniquely linked to biology in the universe.

physics.bio-ph↗
arXiv2023

Power efficient ReLU design for neuromorphic computing using spin Hall effect

We demonstrate a magnetic tunnel junction injected with spin Hall current to exhibit linear rotation of magnetization of the free-ferromagnet using only the spin current. Using the linear resistance change of the MTJ, we devise a circuit for the rectified linear activation (ReLU) function of the artificial neuron. We explore the role of different spin Hall effect (SHE) heavy metal layers on the power consumption of the ReLU circuit. We benchmark the power consumption of the ReLU circuit with different SHE layers by defining a new parameter called the spin Hall power factor. It combines the spin Hall angle, resistivity, and thickness of the heavy metal layer, which translates to the power consumption of the different SHE layers during spin-orbit switching/rotation of the free FM. We employ a hybrid spintronics-CMOS simulation framework that couples Keldysh non-equilibrium Green's function formalism with Landau-Lifshitz-Gilbert-Slonzewski equations and the HSPICE circuit simulator to account for diverse physics of spin-transport and the CMOS elements in our proposed ReLU design. We also demonstrate the robustness of the proposed ReLU circuit against thermal noise and non-trivial power-error trade-off that enables the use of an unstable free-ferromagnet for energy-efficient design. Using the proposed circuit, we evaluate the performance of the convolutional neural network for MNIST datasets and demonstrate comparable classification accuracies to the ideal ReLU with an energy consumption of 75 $pJ$ per sample.

cond-mat.mes-hall↗
arXiv2023

Multimodal N-of-1 trials: A Novel Personalized Healthcare Design

N-of-1 trials aim to estimate treatment effects on the individual level and can be applied to personalize a wide range of physical and digital interventions in mHealth. In this study, we propose and apply a framework for multimodal N-of-1 trials in order to allow the inclusion of health outcomes assessed through images, audio or videos. We illustrate the framework in a series of N-of-1 trials that investigate the effect of acne creams on acne severity assessed through pictures. For the analysis, we compare an expert-based manual labelling approach with different deep learning-based pipelines where in a first step, we train and fine-tune convolutional neural networks (CNN) on the images. Then, we use a linear mixed model on the scores obtained in the first step in order to test the effectiveness of the treatment. The results show that the CNN-based test on the images provides a similar conclusion as tests based on manual expert ratings of the images, and identifies a treatment effect in one individual. This illustrates that multimodal N-of-1 trials can provide a powerful way to identify individual treatment effects and can enable large-scale studies of a large variety of health outcomes that can be actively and passively assessed using technological advances in order to personalized health interventions.

stat.AP↗
arXiv2023

Orthogonal Spin Current Injected Magnetic Tunnel Junction for Convolutional Neural Networks

We propose that a spin Hall effect driven magnetic tunnel junction device can be engineered to provide a continuous change in the resistance across it when injected with orthogonal spin currents. Using this concept, we develop a hybrid device-circuit simulation platform to design a network that realizes multiple functionalities of a convolutional neural network. At the atomistic level, we use the Keldysh non-equilibrium Green's function technique that is coupled self-consistently with the stochastic Landau-Lifshitz-Gilbert-Slonczewski equations, which in turn is coupled with the HSPICE circuit simulator. We demonstrate the simultaneous functionality of the proposed network to evaluate the rectified linear unit and max-pooling functionalities. We present a detailed power and error analysis of the designed network against the thermal stability factor of the free ferromagnets. Our results show that there exists a non-trivial power-error trade-off in the proposed network, which enables an energy-efficient network design based on unstable free ferromagnets with reliable outputs. The static power for the proposed ReLU circuit is $0.56μW$ and whereas the energy cost of a nine-input rectified linear unit-max-pooling network with an unstable free ferromagnet($Δ=15$) is $3.4pJ$ in the worst-case scenario. We also rationalize the magnetization stability of the proposed device by analyzing the vanishing torque gradient points.

cond-mat.mes-hall↗