BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Nirmalya Ghosh

Nirmalya Ghosh contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
physics.opticsComputer VisionMachine Learningeess.IVphysics.soc-phquant-ph

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
13works
0followers
6topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

13 published item(s)

preprint2026arXiv

Geometry Driven Spin Polarization Effects in Waveguiding Plasmonic Crystal

We investigated the modulation in the polarization-dependent optical behaviour of the waveguiding plasmonic crystal by varying the illumination and detection geometry. We employed the finite element method-based COMSOL simulation and Jones-Mueller formalisms to probe the variations in optical parameters by systematically varying the angle of incidence and azimuthal angle of the incident plane wave source, thereby quantifying the variations in the polarization anisotropy parameters. The enhancement of optical properties at various resonances and the tunability of the hybridized modes are discussed. Importantly, our study reveals that despite the system being perfectly inversion-symmetric and achiral, it exhibits circular polarization anisotropy effects for non-zero finite azimuthal angles, manifested in the off-block-diagonal elements of the constructed Mueller matrices. The origin of this intriguing circular anisotropy effect is unravelled using the sequential linear birefringence and linear diattenuation effects arising from geometrical polarization transformation. Its implications in the spin-orbit interaction of light in the plasmonic crystal system are discussed.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Manipulating the transverse spin angular momentum and Belinfante momentum of spin-polarized light by a tilted stratified medium in optical tweezers

In the recent past, optical tweezers incorporating a stratified medium have been exploited to generate complex translational and rotational dynamics in mesoscopic particles due to the coupling between the spin and orbital angular momentum of the light, generated as a consequence of the tight focusing of light by a high numerical aperture objective lens into the stratified medium. Here, we consider an optical tweezers system with a tilted stratified medium (direction of stratification at an angle with the axis of the incident beam), and show that for input circularly polarized Gaussian beams, the resulting spin-orbit interaction deeply influences the generation of transverse spin angular momentum (TSAM) and Belinfante momentum of light, and allows additional control on their magnitude. Importantly, the TSAM generated in our system consists of both the orthogonal components, which is in sharp contrast to the case of evanescent waves and surface plasmons, where only one of the TSAM components are generated. The broken symmetry due to the tilt ensures that, depending upon the helicity of the input beam, the magnitude of the mutually orthogonal components of the TSAM depend entirely on the tilt angle. This may prove to be an effective handle in exotic spin-controlled manipulation of particles in experiments.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Momentum domain polarization probing of forward and inverse spin Hall effect of leaky modes in plasmonic crystals

Simultaneous manifestation of both forward and inverse photonic Spin Hall effect in geometrically tailored anisotropic waveguided plasmonic crystal system is observed through the excitation of the leaky hybridized quasiguided modes. The quasiguided modes of the plasmonic crystal manifested in the far-field momentum (Fourier) domain as circular ring-like intensity distributions, and the different spin orbit interaction (SOI) effects revealed their exclusive signature as polarization-dependent azimuthal intensity lobes on top of it. Using a darkfield Fourier domain polarization Mueller matrix platform, we have observed input spin (circular polarization)- dependent trajectory of the leaky quasiguided modes in the momentum domain (forward spin Hall effect) and its reciprocal effect as the wave vector-controlled spin selection of quasiguided modes (inverse spin Hall effect). These effects are separately manifested in characteristic Mueller matrix elements enabling their interpretation as geometrical circular anisotropy effects. Resonance-enabled enhancement and control of these effects are also demonstrated by exploiting the spectral Fano-type resonance. The far-field manifestation of spin-directional excitation of leaky quasiguided modes, their unique interpretation through momentum domain Mueller matrix, regulation and control of the SOI effects in plasmonic-photonic crystal systems, opens up exciting avenues in spin-orbit photonic research.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Non-separable Optical Beam Shifts and Emergence of Position-position Classical entanglement

Under the introduction of any interface in its trajectory, an optical beam experiences polarization-dependent deflections in the longitudinal and transverse directions with respect to the plane of incidence. The physics of such optical beam shifts is connected to profound universal wave phenomena governed by the fine interference effects of wave packets and has opened up avenues towards metrological applications. Here, we reveal the inherent non-separability of the longitudinal and transverse beam shifts by considering a rather simple case of a partially reflecting Gaussian laser beam from a dielectric interface. This non-separability appears substantially in some particular regions in the corresponding parameter space. We further show that such non-separability manifests as a position-position classically entangled state of light. The tunability of the related experimental parameters offers control over the degree of entanglement. Uncovering of the inherent non-separability of the two types of beam shifts is expected to enrich the physical origin of this fundamental effect, impact the understanding of numerous analogous effects, and might find useful applications by exploiting the position-position-polarization classical entanglement in a fundamental Gaussian beam.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Tunable giant two-dimensional optical beam shift from a tilted linear polarizer

We demonstrate an intriguing giant optical beam shift in a tilted polarizer system analogous to the Imbert-Fedorov shift in partial reflection around the Brewster's angle of incidence. We explain this giant shift using a generalized theoretical treatment for beam shift in a tilted uniaxial anisotropic material incorporating the three-dimensional orientation of its optic axis in the laboratory frame. We further demonstrate regulated control and tunability of both the magnitude and direction of the shifts by wisely tuning the input polarization and the corresponding orientation angles. These findings open up the possibility of precise micron-level beam steering using a simple linear polarizer.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Interferometric weak value of polarization observable and differential Jones matrix algebra

The quantification of simultaneously present weak polarization anisotropy effects are of practical interest from polarimetric and metrological perspective. Recently, in Phys. Rev. A 103, 053518, we experimentally demonstrated a classical analog of post-selected quantum weak measurement through optical interferometry to amplify all possible weak polarization anisotropy effects individually. Here, we propose an extension of this interferometric framework to quantify simultaneously present polarization anisotropy effects. Moreover, a clear correspondence of differential Jones matrix approach with the present scheme is indicated. The proposed scheme enables the measurement of differential Jones matrices through characteristic Stokes vector elements. Our proposal leads to a new class of polarimeter for experimental detection of differential Jones matrix of non-depolarizing anisotropic medium exhibiting simultaneous multiple polarimetric effects of tiny magnitude.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

A Systematic Search over Deep Convolutional Neural Network Architectures for Screening Chest Radiographs

Chest radiographs are primarily employed for the screening of pulmonary and cardio-/thoracic conditions. Being undertaken at primary healthcare centers, they require the presence of an on-premise reporting Radiologist, which is a challenge in low and middle income countries. This has inspired the development of machine learning based automation of the screening process. While recent efforts demonstrate a performance benchmark using an ensemble of deep convolutional neural networks (CNN), our systematic search over multiple standard CNN architectures identified single candidate CNN models whose classification performances were found to be at par with ensembles. Over 63 experiments spanning 400 hours, executed on a 11:3 FP32 TensorTFLOPS compute system, we found the Xception and ResNet-18 architectures to be consistent performers in identifying co-existing disease conditions with an average AUC of 0.87 across nine pathologies. We conclude on the reliability of the models by assessing their saliency maps generated using the randomized input sampling for explanation (RISE) method and qualitatively validating them against manual annotations locally sourced from an experienced Radiologist. We also draw a critical note on the limitations of the publicly available CheXpert dataset primarily on account of disparity in class distribution in training vs. testing sets, and unavailability of sufficient samples for few classes, which hampers quantitative reporting due to sample insufficiency.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

A Two-Stage Multiple Instance Learning Framework for the Detection of Breast Cancer in Mammograms

Mammograms are commonly employed in the large scale screening of breast cancer which is primarily characterized by the presence of malignant masses. However, automated image-level detection of malignancy is a challenging task given the small size of the mass regions and difficulty in discriminating between malignant, benign mass and healthy dense fibro-glandular tissue. To address these issues, we explore a two-stage Multiple Instance Learning (MIL) framework. A Convolutional Neural Network (CNN) is trained in the first stage to extract local candidate patches in the mammograms that may contain either a benign or malignant mass. The second stage employs a MIL strategy for an image level benign vs. malignant classification. A global image-level feature is computed as a weighted average of patch-level features learned using a CNN. Our method performed well on the task of localization of masses with an average Precision/Recall of 0.76/0.80 and acheived an average AUC of 0.91 on the imagelevel classification task using a five-fold cross-validation on the INbreast dataset. Restricting the MIL only to the candidate patches extracted in Stage 1 led to a significant improvement in classification performance in comparison to a dense extraction of patches from the entire mammogram.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Direct observation of the effects of spin dependent momentum of light in optical tweezers

We demonstrate that tight focusing of a circularly polarized Gaussian beam in optical tweezers leads to spin-momentum locking - with the transverse spin angular momentum density being independent of helicity, while the transverse momentum (Poynting vector) becomes helicity dependent. Our theoretical calculations, numerical simulations, and experiments reveal that the presence of a stratified medium in the path of the trapping beam significantly enhances the magnitude of transverse momentum in the radial direction with respect to the beam axis, and likewise, also leads to high off-axial intensity. This overlap allows us to experimentally observe the circular motion of a birefringent particle, trapped off-axis, in response to an input circularly polarized fundamental Gaussian beam carrying no intrinsic orbital angular momentum. The circular motion is dependent on the helicity of the input beam, so that we can identify it to be the signature of the elusive Belinfante spin in propagating light beams obtained in our optical tweezers setup. Our work can be extended to higher-order beams carrying intrinsic orbital angular momentum leading to simple routes of achieving complex particle manipulation using optical tweezers.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Learning Decision Ensemble using a Graph Neural Network for Comorbidity Aware Chest Radiograph Screening

Chest radiographs are primarily employed for the screening of cardio, thoracic and pulmonary conditions. Machine learning based automated solutions are being developed to reduce the burden of routine screening on Radiologists, allowing them to focus on critical cases. While recent efforts demonstrate the use of ensemble of deep convolutional neural networks(CNN), they do not take disease comorbidity into consideration, thus lowering their screening performance. To address this issue, we propose a Graph Neural Network (GNN) based solution to obtain ensemble predictions which models the dependencies between different diseases. A comprehensive evaluation of the proposed method demonstrated its potential by improving the performance over standard ensembling technique across a wide range of ensemble constructions. The best performance was achieved using the GNN ensemble of DenseNet121 with an average AUC of 0.821 across thirteen disease comorbidities.

0 citations0 reviewsNo code linkedOpen access
preprint2019arXiv

Single-shot measurement of space varying polarization state of light through interferometric quantification of geometric phase

Light beam carrying spatially varying state of polarization generates space varying Pancharatnam-Berry geometric phase while propagating through homogeneous anisotropic medium. We show that determination of such space varying geometric phase provides a unique way to quantify the space varying polarization state of light using a single-shot interferometric measurement. We demonstrate this concept in a Mach-Zehnder interferometric arrangement using a linearly polarized reference light beam, where full information on the spatially varying polarization state is successfully recovered by quantifying the space varying geometric phase and the contrast of interference. The proposed method enables single-shot measurement of any space varying polarization state of light from the measured interference pattern with a polarized reference beam. This approach shows considerable potential for instantaneous mapping of complex space varying polarization of light in diverse applications, such as astronomy, biomedical imaging, nanophotonics, etc., where high precision and near real-time measurement of spatial polarization patterns are desirable.

0 citations0 reviewsNo code linkedOpen access
preprint2019arXiv

Unraveling linguistic patterns in dog behaviour

Apparently random events in nature often reveal hidden patterns when analysed using diverse and robust statistical tools. Power-law distributions, for example, project diverse natural phenomenon, ranging from earthquakes1 to heartbeat dynamics2 onto a common platform of statistical self-similarity. A large range of human languages are known to follow a specific regime of power-law distributions, the Zipf-Mandelbrot law, in addition to showing properties like the Pareto principle and Shannon entropy3,4. Animal behaviour in specific contexts have been shown to follow power-law distributions5,6. However, the entire behavioural repertoire of a species has never been analysed for the existence of underlying patterns. Here we show that the frequency-rank data of randomly sighted behaviours at the population level of free-ranging dogs follow a scale-invariant power-law behaviour. While the data does not display Zipfian trends, it obeys the Pareto principle and Shannon entropy rules akin to languages. Interestingly, the data also exhibits robust self-similarity patterns at different scales which we extract using multifractal detrended fluctuation analysis7. The observed multifractal trends suggest that the probability of consecutive occurrence of behaviours of adjacent ranks is much higher than behaviours widely separated in rank. Since we observe such robust trends in random data sets, we hypothesize that the general behavioural repertoire of a species is shaped by a syntax similar to languages. This opens up the prospect of future multifractal and other statistical investigations on true time series of behavioural data to probe the existence of possible long and short-range correlations, and thereby develop predictive models of behaviour.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

Transverse spin and transverse momentum in scattering of plane waves

We study the near field to the far field evolution of spin angular momentum (SAM) density and the Poynting vector of the scattered waves from spherical scatterers. The results show that at the near field, the SAM density and the Poynting vector are dominated by their transverse components. While the former (transverse SAM) is independent of the helicity of the incident circular polarization state, the latter (transverse Poynting vector) depends upon the polarization state. It is further demonstrated that the magnitudes and the spatial extent of the transverse SAM and the transverse momentum components can be controllably enhanced by exploiting the interference of the transverse electric and transverse magnetic scattering modes.

0 citations0 reviewsNo code linkedOpen access