arXiv subjects

Neeraj Gupta

Publications and source records attributed to Neeraj Gupta.

At least 19 recordsLinked to original sources

arXiv2026

Low-Frequency Recombination Lines from Galaxies and AGN over Cosmic Time

Radio recombination lines (RRLs) at low frequencies (<10 GHz) can provide a multi-phase view of interstellar gas in nearby galaxies, absorption-line-systems, and AGN. Hydrogen RRLs arise in fully ionized gas and carbon RRLs trace elusive cold-HI and CO-dark molecular gas. Low frequency RRLs are typically stimulated by the radio continuum and thus may be observable within or against radio bright sources out to cosmological distances (z ~ 6). Although long sought after, RRLs were only recently detected outside of the local universe (z ~ 1; Emig et al., 2020, 2023). Such detections have been made possible by the advancement of wide-bandwidth spectral-line surveys on next-generation low-frequency telescopes. Precursors and pathfinders to the SKA have opened up this field of research and will make significant advancements over the next years by enabling surveys over large source samples. The SKA will provide access to the crucial frequency ranges where RRL line intensity is brightest. Furthermore, multi-band SKA measurements will fully characterize gas physical conditions. Key extragalactic science of low frequency RRLs will focus on (i) the conversion of baryonic material into stars across cosmic time, (ii) the evolution of the ISM and its physical conditions in galaxies, and (iii) how gas drives and inhibits AGN activity.

astro-ph.GA↗

arXiv2026

Tracing Cold Gas in Absorption Across Cosmic Time with the SKA

Observing the 21-cm HI line in absorption provides a powerful means of tracing the cold neutral gas in normal and active galaxies across cosmic time. The frequency coverage and sensitivity of SKAO will allow us to detect HI in absorption from z = 0 to beyond z = 6, enabling the characterisation of the properties of cold gas in and around galaxies at all epochs. This chapter summarises recent advances in absorption-line studies, lessons learned from precursor surveys, and updates the science case presented in Kanekar and Briggs (2004) and Morganti et al. (2015), focusing on the capabilities enabled by the SKA design baseline, Array Assembly 4 (AA4). We expand on these earlier works by presenting new opportunities to simultaneously search for OH 18-cm absorption, an efficient tracer of diffuse molecular gas that complements the atomic gas traced by HI absorption, as well as the need for sub-arcsecond scale spectroscopic imaging and multi-wavelength data from large surveys. These advances will allow SKAO absorption surveys to address key questions surrounding the fuelling and feedback cycles of AGN and the evolution of the cold neutral gas across cosmic time.

astro-ph.GA↗

arXiv2026

HI Galaxy Science with the SKA

This chapter introduces the contributions of the HI galaxy science in this volume reviewing the latest developments and urgent questions in HI galaxy science, providing guiding principles for a layered set of future key science projects. The key science will include: a complete censuses of HI morphologies and kinematics at sub-kpc and 1 km/s resolution within and around galaxies in the nearby Universe; a measurement of the cosmic HI mass density and HI mass function evolution at least up to z~1; an improved understanding of the Universe at z>1, particularly the balance between cold molecular and cool atomic gas. We also provide a view of the synergistic multi-wavelength surveys available in 2028+ in the southern hemisphere. This effort will improve our understanding of the baryon cycle across a significant fraction of the cosmic history, including the processes of gas accretion, consumption and removal as well as AGN and star formation feedback. Based on these science goals, the earlier proposed three-tiered survey strategy remains, but survey parameters and predictions are adjusted according to AA* and AA4 developments. This chapter is an update of the earlier "Advancing Astrophysics with the Square Kilometre Array" chapter 'HI Science with the SKA' by Staveley-Smith & Oosterloo.

astro-ph.GA↗

arXiv2026

Advancing Astrophysics with the SKA II

Advancing Astrophysics with the SKA II (AASKAII), written by our science community, outlines the transformative scientific advances that will be enabled by the SKA telescopes. In the decade since the publication of the previous edition, telescope designs have matured, construction has commenced, and the SKA Organisation has evolved into the SKA Observatory (SKAO). At the same time, observations from SKA precursor and pathfinder telescopes have provided new insights into longstanding scientific challenges while revealing entirely new phenomena. Published in advance of the first science verification campaign for the SKA Observatory, this volume looks ahead to the coming decades of discovery and innovation in radio astronomy. AASKAII spans the broad range of scientific research enabled by the SKA telescopes, SKA-Mid and SKA-Low. The contributions are organised into six thematic categories according to their scientific focus. The opening section presents overview chapters from the SKA Science Working Groups, around which our community is organised. Each overview provides the broader context that connects the contributions in this volume to the key scientific questions being pursued by their respective communities.

astro-ph.IM↗

arXiv2026

Towards Robust Universal Perturbation Attacks: A Float-Coded, Penalty-Driven Evolutionary Approach

Universal adversarial perturbations (UAPs) have garnered significant attention due to their ability to undermine deep neural networks across multiple inputs using a single noise pattern. Evolutionary algorithms offer a promising approach to generating such perturbations due to their ability to navigate non-convex, gradient-free landscapes. In this work, we introduce a float-coded, penalty-driven single-objective evolutionary framework for UAP generation that achieves lower visibility perturbations while enhancing attack success rates. Our approach leverages continuous gene representations aligned with contemporary deep learning scales, incorporates dynamic evolutionary operators with adaptive scheduling, and utilizes a modular PyTorch implementation for seamless integration with modern architectures. Additionally, we ensure the universality of the generated perturbations by testing across diverse models and by periodically switching batches to prevent overfitting. Experimental results on the ImageNet dataset demonstrate that our framework consistently produces perturbations with smaller norms, higher misclassification effectiveness, and faster convergence compared to existing evolutionary-based methods. These findings highlight the robustness and scalability of our approach for universal adversarial attacks across various deep learning architectures.

cs.LG↗

arXiv2025

H I Properties of Field Galaxies at $\boldsymbol{z\approx 0.2}$-0.6: Insights into Declining Cosmic Star Formation

We report statistically significant detection of H I 21-cm emission from intermediate-redshift ($z\approx0.2$-0.6) galaxies. By leveraging multi-sightline galaxy survey data from the Cosmic Ultraviolet Baryon Survey (CUBS) and deep radio observations from the MeerKAT Absorption Line Survey (MALS), we have established a sample of $\approx6000$ spectroscopically identified galaxies in 11 distinct fields to constrain the neutral gas content at intermediate redshifts. The galaxies sample a broad range in stellar mass -- $8\lesssim\log{M_\rm{star}/\rm{M}_\odot}\lesssim11$ with a median of $\langle\log{M_\rm{star}/\rm{M}_\odot}\rangle_\rm{med}\approx10$ -- and a wide range in redshift -- $0.24\lesssim z\lesssim0.63$ with a median of $\langle z\rangle_\rm{med}=0.44$. Our detected emission-line signal exceeds $4\,\sigma$ significance in the stacked spectra of all subsamples, and the observed total H I 21-cm line flux translates to a H I mass $M_\rm{H\;I}\approx10^{10}\rm{M}_\odot$. We find a high H I-to-stellar mass ratio of $M_\mathrm{H\;I}/M_\rm{star}\approx6$ for low-mass galaxies with $\langle\log{M_\rm{star}/\rm{M}_\odot}\rangle \approx9.3$ ($>3.7\,\sigma$). For galaxies with $\langle\log{M_\rm{star}/\rm{M}_\odot}\rangle\approx10.6$, we find $M_\mathrm{H\;I}/M_\rm{star}\approx0.3$ ($>4.7\,\sigma$). Additionally, the redshift evolution of H I mass in both low- and high-mass field galaxies, inferred from the stacked emission-line signal, aligns well with the expectation from the cosmic star formation history. This suggests that the overall decline in the cosmic star formation activity across the general galaxy population may be connected to a decreasing supply of neutral hydrogen. Finally, our analysis has revealed significant 21-cm signals at distances greater than 75 kpc from these intermediate-redshift galaxies, indicating a substantial reservoir of H I gas in their extended surroundings.

astro-ph.GA↗

arXiv2025

MedChat: A Multi-Agent Framework for Multimodal Diagnosis with Large Language Models

The integration of deep learning-based glaucoma detection with large language models (LLMs) presents an automated strategy to mitigate ophthalmologist shortages and improve clinical reporting efficiency. However, applying general LLMs to medical imaging remains challenging due to hallucinations, limited interpretability, and insufficient domain-specific medical knowledge, which can potentially reduce clinical accuracy. Although recent approaches combining imaging models with LLM reasoning have improved reporting, they typically rely on a single generalist agent, restricting their capacity to emulate the diverse and complex reasoning found in multidisciplinary medical teams. To address these limitations, we propose MedChat, a multi-agent diagnostic framework and platform that combines specialized vision models with multiple role-specific LLM agents, all coordinated by a director agent. This design enhances reliability, reduces hallucination risk, and enables interactive diagnostic reporting through an interface tailored for clinical review and educational use. Code available at https://github.com/Purdue-M2/MedChat.

cs.MA↗

arXiv2025

Quasar radiation transforms the gas in a merging companion galaxy

Quasars, powered by gas accretion onto supermassive black holes, rank among the most energetic objects of the Universe. While they are thought to be ignited by galaxy mergers and affect the surrounding gas, observational constraints on both processes remain scarce. Here we unveil a major merging system at redshift $z \approx 2.7$, and demonstrate that radiation from the quasar in one galaxy directly alters the gas properties in the other galaxy. Our findings reveal that the galaxies, with centroids separated by only a few kiloparsecs and approaching each other at speed $\approx550\,$km$\,$s$^{-1}$, are massive, form stars, and contain a substantial molecular mass. Yet, dusty molecular gas seen in absorption against the quasar nucleus is highly excited and confined within cloudlets with densities $\sim 10^5$ - $10^6$ cm$^{-3}$ and sizes $<$0.02 pc, several orders of magnitude more compact than those observed in intervening (non-quasar) environments. This is also approximately 10$^5$ times smaller than currently resolvable through molecular-line emission at high redshifts. We infer that, wherever exposed to the quasar radiation, molecular gas is disrupted, leaving behind surviving dense clouds too small to give birth to new stars. Our results not only underscore the role of major galaxy mergers in triggering quasar activity, but also reveal localized negative feedback as a profound alteration of internal gas structure which likely hampers star formation.

astro-ph.GA↗

arXiv2025

New constraints on the values of the fundamental constants at a look-back time of 7.3 Gyr

The study of redshifted spectral lines can provide a measure of the fundamental constants over large look-back times. Current grand unified theories predict an evolution in these constants and astronomical observations offer the only experimental measure of the values of the constants over large timescales. Of particular interest are the dimensionless constants: the fine structure constant ($\alpha$), the proton-electron mass ratio ($\mu$), and the proton g-factor ($g_p$), since these do not require a "standard meterstick". Here we present a re-analysis of the 18 cm hydroxyl (OH) lines at $z=0.89$, which were recently detected with the MeerKAT telescope, toward the radio source PKS\,1830-211. Utilizing the previous constraint of $\Delta\mu/\mu=(-1.8\pm1.2)\times10^{-7}$, we obtain $\Delta (\alpha g_p^{0.27})/(\alpha g_p^{0.27})\lesssim5.7\times10^{-5}$, $\Delta \alpha/\alpha\lesssim2.3\times10^{-3}$ , and $\Delta g_p/g_p\lesssim7.9\times10^{-3}$. These new constraints are consistent with no evolution over a look-back time of 7.3 Gyr and provide another valuable data point in the putative evolution of the constants.

astro-ph.CO↗

arXiv2024

Fermi Unassociated Sources in the MeerKAT Absorption Line Survey

Over 2000 Gamma ray sources identified by the Large Area Telescope (LAT) on NASA's Fermi Gamma-ray Space Telescope are considered unassociated, meaning that they have no known counterparts in any other frequency regime. We have carried out an image-based search for steep spectrum radio sources, with in-band spectral index less than -1.4, within the error regions of Fermi unassociated sources using 1 to 1.4 GHz radio data from the MeerKAT Absorption Line Survey (MALS) Data Release. MALS DR1 with a median rms noise of 22 to 25 microJy and 735,649 sources is a significant advance over past image-based searches with improvements in sensitivity, resolution and bandwidth. Steep spectrum candidates were identified using a combination of in-band spectral indices from MALS and existing radio surveys. We developed an optical and infrared source classification scheme in order to distinguish between galactic pulsars and radio galaxies. In total, we identify nine pulsar candidates towards six Fermi sources that are worthy of follow-up for pulsation searches. We also report 41 steep spectrum radio galaxy candidates that may be of interest in searches for high-redshift radio galaxies. We show that MALS due to its excellent continuum sensitivity can detect 80 percent of the known pulsar population. This exhibits the promise of identifying exotic pulsar candidates with future image-based surveys with the Square Kilometre and its precursors.

astro-ph.HE↗

arXiv2024

Detecting Multimedia Generated by Large AI Models: A Survey

The rapid advancement of Large AI Models (LAIMs), particularly diffusion models and large language models, has marked a new era where AI-generated multimedia is increasingly integrated into various aspects of daily life. Although beneficial in numerous fields, this content presents significant risks, including potential misuse, societal disruptions, and ethical concerns. Consequently, detecting multimedia generated by LAIMs has become crucial, with a marked rise in related research. Despite this, there remains a notable gap in systematic surveys that focus specifically on detecting LAIM-generated multimedia. Addressing this, we provide the first survey to comprehensively cover existing research on detecting multimedia (such as text, images, videos, audio, and multimodal content) created by LAIMs. Specifically, we introduce a novel taxonomy for detection methods, categorized by media modality, and aligned with two perspectives: pure detection (aiming to enhance detection performance) and beyond detection (adding attributes like generalizability, robustness, and interpretability to detectors). Additionally, we have presented a brief overview of generation mechanisms, public datasets, online detection tools, and evaluation metrics to provide a valuable resource for researchers and practitioners in this field. Most importantly, we offer a focused analysis from a social media perspective to highlight their broader societal impact. Furthermore, we identify current challenges in detection and propose directions for future research that address unexplored, ongoing, and emerging issues in detecting multimedia generated by LAIMs. Our aim for this survey is to fill an academic gap and contribute to global AI security efforts, helping to ensure the integrity of information in the digital realm. The project link is https://github.com/Purdue-M2/Detect-LAIM-generated-Multimedia-Survey.

cs.MM↗

arXiv2023

Cold molecules in HI 21cm absorbers across redshifts 0.1-4

Absorption lines at high redshift in front of quasars are rare in the mm domain. Only five associated and five intervening systems have been reported in the literature. These bring very useful information complementary to emission lines, for instance, to distinguish between inflows and outflows. They are also good candidates to study the variations of the fundamental constants. We report here the search for molecules in emission and absorption in front of a sample of 30 targets, comprising 16 associated and 14 intervening HI 21-cm absorbers. The observations have been done with the IRAM-30m telescope, simultaneously at 3mm and 2mm, exploring CO ladder and HCO+ lines. Eight targets have been detected in emission, of which five are new. Their molecular gas masses range from 10^9 to 7 10^11 Mo. We also report four new detections in absorption. Two of the associated CO absorption line detections at high redshift (z=1.211 and 1.275) resulted from the high spatial resolution follow-up with NOEMA. The disparity between the mm molecular and HI 21-cm absorption lines for these and another intervening system detected in HNC at z = 1.275, is attributable to radio and mm sight lines tracing different media. Comparing HI and H2 in the 14 known high redshift molecular absorbers, associated HI absorption lines are broad, with multiple components and the molecular absorption corresponds to the broader and weaker 21-cm absorption component. This indicates two distinct phases: one near galaxy centers with a larger CO-to-HI abundance ratio, and another with lower molecular abundance in the outer regions of the galaxy. The comparison of interferometric and single dish observations shows that the detection of absorption requires sufficient spatial resolution to overcome the dilution by emission, and will be an important criterion for mm follow-up of 21-cm absorbers from ongoing large-scale surveys.

astro-ph.GA↗

arXiv2023

Discovery of Hydrogen Radio Recombination Lines at z=0.89 towards PKS 1830-211

We report the detection of stimulated hydrogen radio recombination line (RRL) emission from ionized gas in a $z=0.89$ galaxy using 580--1670 MHz observations from the MeerKAT Absorption Line Survey (MALS). The RRL emission originates in a galaxy that intercepts and strongly lenses the radio blazar PKS 1830-211 ($z=2.5$). This is the second detection of RRLs outside of the local universe and the first clearly associated with hydrogen. We detect effective H144$\alpha$ (and H163$\alpha$) transitions at observed frequencies of 1156 (798) MHz by stacking 17 (27) RRLs with 21$\sigma$ (14$\sigma$) significance. The RRL emission contains two main velocity components and is coincident in velocity with HI 21 cm and OH 18 cm absorption. We use the RRL spectral line energy distribution and a Bayesian analysis to constrain the density ($n_e$) and the volume-averaged pathlength ($\ell$) of the ionized gas. We determine $\log( n_e ) = 2.0_{-0.7}^{+1.0}$ cm$^{-3}$ and $\log( \ell ) = -0.7_{-1.1}^{+1.1}$ pc towards the north east (NE) lensed image, likely tracing the diffuse thermal phase of the ionized ISM in a thin disk. Towards the south west (SW) lensed image, we determine $\log( n_e ) = 3.2_{-1.0}^{+0.4}$ cm$^{-3}$ and $\log( \ell ) = -2.7_{-0.2}^{+1.8}$ pc, tracing gas that is more reminiscent of H II regions. We estimate a star formation (surface density) rate of $\Sigma_{\mathrm{SFR}} \sim 0.6$ M$_{\odot}$ yr$^{-1}$ kpc$^{-2}$ or SFR $\sim 50$ M$_{\odot}$ yr$^{-1}$, consistent with a star-forming main sequence galaxy of $M_{\star} \sim 10^{11}$ M$_{\odot}$. The discovery presented here opens up the possibility of studying ionized gas at high redshifts using RRL observations from current and future (e.g., SKA and ngVLA) radio facilities.

astro-ph.GA↗

arXiv2022

Emergence of a new HI 21-cm absorption component at z~1.1726 towards the gamma-ray blazar PKS~2355-106

We report the emergence of a new HI 21-cm absorption at z_abs = 1.172635 in the damped Lyman-alpha absorber (DLA) towards the gamma-ray blazar PKS 2355-106 (z_em~1.639) using science verification observations (June 2020) from the MeerKAT Absorption Line Survey (MALS). Since 2006, this DLA is known to show a narrow HI 21-cm absorption at z_abs = 1.173019 coinciding with a distinct metal absorption line component. We do not detect significant HI 21-cm optical depth variations from this known HI component. A high resolution optical spectrum (August 2010) shows a distinct Mg I absorption at the redshift of the new HI 21-cm absorber. However, this component is not evident in the profiles of singly ionized species. We measure the metallicity ([Zn/H] = -(0.77\pm0.11) and [Si/H]= -(0.96\pm0.11)) and depletion ([Fe/Zn] = -(0.63\pm0.16)) for the full system. Using the apparent column density profiles of Si II, Fe II and Mg I we show that the depletion and the N(Mg I)/N(Si II) column density ratio systematically vary across the velocity range. The region with high depletion tends to have slightly larger N(Mg I)/N(Si II) ratio. The two HI 21-cm absorbers belong to this velocity range. The emergence of z_abs = 1.172635 can be understood if there is a large optical depth gradient over a length scale of ~0.35 pc. However, the gas producing the z_abs = 1.173019 component must be nearly uniform over the same scale. Systematic uncertainties introduced by the absorption line variability has to be accounted for in experiments measuring the variations of fundamental constants and cosmic acceleration even when the radio emission is apparently compact as in PKS 2355-106.

astro-ph.GA↗

arXiv2022

Discovery of a damped Ly$\alpha$ absorber originating in a spectacular interacting dwarf galaxy pair at $z = 0.026$

We present the discovery of neutral gas detected in both damped Ly$\alpha$ absorption (DLA) and HI 21-cm emission outside of the stellar body of a galaxy, the first such detection in the literature. A joint analysis between the Cosmic Ultraviolet Baryon Survey and the MeerKAT Absorption Line Survey reveals an HI bridge connecting two interacting dwarf galaxies (log$(M_{\text{star}}/\text{M}_{\odot}) = 8.5 \pm 0.2$) that host a $z = 0.026$ DLA with log[$N$(HI)/cm$^{-2}$]$ = 20.60 \pm 0.05$ toward the QSO J2339-5523 ($z_{\text{QSO}} = 1.35$). At impact parameters of $d = 6$ and $33$ kpc, the dwarf galaxies have no companions more luminous than $\approx 0.05L_{*}$ within at least $\Delta v = \pm 300$ km s$^{-1}$ and $d \approx 350$ kpc. HI 21-cm emission is spatially coincident with the DLA at the 2-3$\sigma$ level per spectral channel over several adjacent beams. However, HI 21-cm absorption is not detected against the radio-bright QSO; if the background UV and radio sources are spatially aligned, the gas is either warm or clumpy (with spin temperature to covering factor ratio $T_{s}/f_{c} > 1880$ K). VLT-MUSE observations demonstrate that the $\alpha$-element abundance of the ionized ISM is consistent with the DLA ($\approx 10$% solar), suggesting that the neutral gas envelope is perturbed ISM gas. This study showcases the impact of dwarf-dwarf interactions on the physical and chemical state of neutral gas outside of star-forming regions. In the SKA era, joint UV and HI 21-cm analyses will be critical for connecting the cosmic neutral gas content to galaxy environments.

astro-ph.GA↗

arXiv2021

A GMRT Narrowband vs. Wideband Analysis of the ACT-CLJ0034.4+0225 Field Selected from the ACTPol Cluster Sample

Low frequency radio observations of galaxy clusters are a useful probe of the non-thermal intracluster medium (ICM), through observations of diffuse radio emission such as radio halos and relics. Current formation theories cannot fully account for some of the observed properties of this emission. In this study, we focus on the development of interferometric techniques for extracting extended, faint diffuse emissions in the presence of bright, compact sources in wide-field and broadband continuum imaging data. We aim to apply these techniques to the study of radio halos, relics and radio mini-halos using a uniformly selected and complete sample of galaxy clusters selected via the Sunyaev-Zel'dovich (SZ) effect by the Atacama Cosmology Telescope (ACT) project, and its polarimetric extension (ACTPol). We use the upgraded Giant Metrewave Radio Telescope (uGMRT) for targeted radio observations of a sample of 40 clusters. We present an overview of our sample, confirm the detection of a radio halo in ACT-CL J0034.4+0225, and compare the narrowband and wideband analysis results for this cluster. Due to the complexity of the ACT-CL J0034.4+0225 field, we use three pipelines to process the wideband data. We conclude that the experimental SPAM wideband pipeline produces the best results for this particular field. However, due to the severe artefacts in the field, further analysis is required to improve the image quality.

astro-ph.CO↗

arXiv2021

Damped Ly-alpha Absorbers in Star-forming Galaxies at z < 0.15 Detected with the Hubble Space Telescope and Implications for Galaxy Evolution

We report {\it HST} COS spectroscopy of 10 quasars with foreground star-forming galaxies at 0.02$<$$z$$<$ 0.14 within impact parameters of $\sim$1-7 kpc. We detect damped/sub-damped Ly$\alpha$ absorption in 100$\%$ of cases where no higher-redshift Lyman-limit systems extinguish the flux at the expected wavelength of Ly$\alpha$ absorption, obtaining the largest targeted sample of DLA/sub-DLAs in low-redshift galaxies. We present absorption measurements of neutral hydrogen and metals. Additionally, we present GBT 21-cm emission measurements for 5 of the galaxies (including 2 detections). Combining our sample with the literature, we construct a sample of 115 galaxies associated with DLA/sub-DLAs spanning 0$<$$z$$<$4.4, and examine trends between gas and stellar properties, and with redshift. The H~I column density is anti-correlated with impact parameter and stellar mass. More massive galaxies appear to have gas-rich regions out to larger distances. The specific SFR (sSFR) of absorbing galaxies increases with redshift and decreases with $M^{\ast}$, consistent with evolution of the star-formation main sequence (SFMS). However, $\sim$20$\%$ of absorbing galaxies lie below the SFMS, indicating that some DLA/sub-DLAs trace galaxies with longer-than-typical gas-depletion time-scales. Most DLA/sub-DLA galaxies with 21-cm emission have higher H I masses than typical galaxies with comparable $M^{\ast}$. High $M_{\rm H I}/M^{\ast}$ ratios and high sSFRs in DLA/sub-DLA galaxies with $M^{\ast}$$<$$10^{9}$$M_{\odot}$ suggest these galaxies may be gas-rich because of recent gas accretion rather than inefficient star formation. Our study demonstrates the power of absorption and emission studies of DLA/sub-DLA galaxies for extending galaxy-evolution studies to previously under-explored regimes of low $M^{\ast}$ and low SFR.

astro-ph.GA↗

arXiv2021

FREGAN : an application of generative adversarial networks in enhancing the frame rate of videos

A digital video is a collection of individual frames, while streaming the video the scene utilized the time slice for each frame. High refresh rate and high frame rate is the demand of all high technology applications. The action tracking in videos becomes easier and motion becomes smoother in gaming applications due to the high refresh rate. It provides a faster response because of less time in between each frame that is displayed on the screen. FREGAN (Frame Rate Enhancement Generative Adversarial Network) model has been proposed, which predicts future frames of a video sequence based on a sequence of past frames. In this paper, we investigated the GAN model and proposed FREGAN for the enhancement of frame rate in videos. We have utilized Huber loss as a loss function in the proposed FREGAN. It provided excellent results in super-resolution and we have tried to reciprocate that performance in the application of frame rate enhancement. We have validated the effectiveness of the proposed model on the standard datasets (UCF101 and RFree500). The experimental outcomes illustrate that the proposed model has a Peak signal-to-noise ratio (PSNR) of 34.94 and a Structural Similarity Index (SSIM) of 0.95.

cs.CV↗