Search

The key to detecting neutral hydrogen during the epoch of reionization (EoR) is to separate the cosmological signal from the dominating foreground radiation. We developed direct optimal mapping (DOM) to map interferometric visibilities; it contains only linear operations, with full knowledge of point spread functions from visibilities to images. Here, we demonstrate a fast Fourier transform-based image power spectrum and its window functions computed from the DOM images. We use noiseless simulation, based on the Hydrogen Epoch of Reionization Array Phase I configuration, to study the image power spectrum properties. The window functions show $<10^{-11}$ of the integrated power leaks from the foreground-dominated region into the EoR window; the 2D and 1D power spectra also verify the separation between the foregrounds and the EoR., Comment: Published in ApJ

We consider essential self-adjointness on the space $C_0^{\infty}((0,\infty))$ of even order, strongly singular, homogeneous differential operators associated with differential expressions of the type \[ \tau_{2n}(c) = (-1)^n \frac{d^{2n}}{d x^{2n}} + \frac{c}{x^{2n}}, \quad x > 0, \; n \in \mathbb{N}, \; c \in \mathbb{R}, \] in $L^2((0,\infty);dx)$. While the special case $n=1$ is classical and it is well-known that $\tau_2(c)\big|_{C_0^\infty((0,\infty))}$ is essentially self-adjoint if and only if $c \geq 3/4$, the case $n \in \mathbb{N}$, $n \geq 2$, is far from obvious. In particular, it is not at all clear from the outset that \[ \text{ there exists } c_n \in \mathbb{R}, \, n \in \mathbb{N}, \text{ such that } \tau_{2n}(c)\big|_{C_0^\infty((0,\infty))} \, \text{ is essentially self-adjoint if and only if } c \geq c_n. \tag{*}\label{0.1} \] As one of the principal results of this paper we indeed establish the existence of $c_n$, satisfying $c_n \geq (4n-1)!!\big/2^{2n}$, such that property \eqref{0.1} holds. In sharp contrast to the analogous lower semiboundedness question, \[ \text{ for which values of } c \, \text{\it is } \tau_{2n}(c)\big|_{C_0^{\infty}((0,\infty))} \, \text{ bounded from below?}, \] which permits the sharp (and explicit) answer $c \geq [(2n -1)!!]^{2}\big/2^{2n}$, $n \in \mathbb{N}$, the answer for \eqref{0.1} is surprisingly complex and involves various aspects of the geometry and analytical theory of polynomials. For completeness we record explicitly, \[ c_1 = 3/4, \quad c_2= 45, \quad c_3 = 2240 \big(214+7 \sqrt{1009}\,\big)\big/27, \] and remark that $c_n$ is the root of a polynomial of degree $n-1$. We demonstrate that for $n=6,7$, $c_n$ are algebraic numbers not expressible as radicals over $\mathbb{Q}$ (and conjecture this is in fact true for general $n \geq 6$)., Comment: 36 pages

We provide an elementary derivation of the Bessel analog of the celebrated Riesz composition formula and use the former to effortlessly derive the latter., Comment: 21 pages

Large sky spectroscopic surveys have reached the scale of photometric surveys in terms of sample sizes and data complexity. These huge datasets require efficient, accurate, and flexible automated tools for data analysis and science exploitation. We present the Galaxy Spectra Network/GaSNet-II, a supervised multi-network deep learning tool for spectra classification and redshift prediction. GaSNet-II can be trained to identify a customized number of classes and optimize the redshift predictions for classified objects in each of them. It also provides redshift errors, using a network-of-networks that reproduces a Monte Carlo test on each spectrum, by randomizing their weight initialization. As a demonstration of the capability of the deep learning pipeline, we use 260k Sloan Digital Sky Survey spectra from Data Release 16, separated into 13 classes including 140k galactic, and 120k extragalactic objects. GaSNet-II achieves 92.4% average classification accuracy over the 13 classes (larger than 90% for the majority of them), and an average redshift error of approximately 0.23% for galaxies and 2.1% for quasars. We further train/test the same pipeline to classify spectra and predict redshifts for a sample of 200k 4MOST mock spectra and 21k publicly released DESI spectra. On 4MOST mock data, we reach 93.4% accuracy in 10-class classification and an average redshift error of 0.55% for galaxies and 0.3% for active galactic nuclei. On DESI data, we reach 96% accuracy in (star/galaxy/quasar only) classification and an average redshift error of 2.8% for galaxies and 4.8% for quasars, despite the small sample size available. GaSNet-II can process ~40k spectra in less than one minute, on a normal Desktop GPU. This makes the pipeline particularly suitable for real-time analyses of Stage-IV survey observations and an ideal tool for feedback loops aimed at night-by-night survey strategy optimization., Comment: 23 pages and 31 figures. The published version of MNRAS

The analogue of Hilbert's tenth problem over $\mathbb{Q}$ asks for an algorithm to decide the existence of rational points in algebraic varieties over this field. This remains as one of the main open problems in the area of undecidability in number theory. Besides the existence of rational points, there is also considerable interest in the problem of effectivity: one asks whether the sought rational points satisfy determined height bounds, often expressed in terms of the height of the coefficients of the equations defining the algebraic varieties under consideration. We show that, in fact, Hilbert's tenth problem over $\mathbb{Q}$ with (finitely many) height comparison conditions is undecidable., Comment: References updated

There is a current increase in the development of "4D" Doppler-capable radar and lidar range sensors that produce 3D point clouds where all points also have information about the radial velocity relative to the sensor. 4D radars in particular are interesting for object perception and navigation in low-visibility conditions (dust, smoke) where lidars and cameras typically fail. With the advent of high-resolution Doppler-capable radars comes the possibility of estimating odometry from single point clouds, foregoing the need for scan registration which is error-prone in feature-sparse field environments. We compare several odometry estimation methods, from direct integration of Doppler/IMU data and Kalman filter sensor fusion to 3D scan-to-scan and scan-to-map registration, on three datasets with data from two recent 4D radars and two IMUs. Surprisingly, our results show that the odometry from Doppler and IMU data alone give similar or better results than 3D point cloud registration. In our experiments, the average position error can be as low as 0.3% over 1.8 and 4.5km trajectories. That allows accurate estimation of 6DOF ego-motion over long distances also in feature-sparse mine environments. These results are useful not least for applications of navigation with resource-constrained robot platforms in feature-sparse and low-visibility conditions such as mining, construction, and search & rescue operations., Comment: Preprint. Submitted to ICRA 2024. 7 pages, 11 figures

For an elliptic curve $E$ defined over the field $\mathbb{C}$ of complex numbers, we classify all translates of elliptic curves in $E^3$ such that the $x$-coordinates satisfy a linear equation. This classification enables us to establish a relation between the rank of finite rank subgroups of $E$ and triples in $E$ whose $x$-coordinates are linearly related. The method of proof integrates complex analytic techniques on elliptic curves with results of Gao, Ge and K\"uhne on Uniform Mordell-Lang Conjecture for subvarieties in abelian varieties., Comment: 8 pages, comments are welcome

While advanced machine learning (ML) models are deployed in numerous real-world applications, previous works demonstrate these models have security and privacy vulnerabilities. Various empirical research has been done in this field. However, most of the experiments are performed on target ML models trained by the security researchers themselves. Due to the high computational resource requirement for training advanced models with complex architectures, researchers generally choose to train a few target models using relatively simple architectures on typical experiment datasets. We argue that to understand ML models' vulnerabilities comprehensively, experiments should be performed on a large set of models trained with various purposes (not just the purpose of evaluating ML attacks and defenses). To this end, we propose using publicly available models with weights from the Internet (public models) for evaluating attacks and defenses on ML models. We establish a database, namely SecurityNet, containing 910 annotated image classification models. We then analyze the effectiveness of several representative attacks/defenses, including model stealing attacks, membership inference attacks, and backdoor detection on these public models. Our evaluation empirically shows the performance of these attacks/defenses can vary significantly on public models compared to self-trained models. We share SecurityNet with the research community. and advocate researchers to perform experiments on public models to better demonstrate their proposed methods' effectiveness in the future., Comment: To appear in the 33rd USENIX Security Symposium, August 2024, Philadelphia, PA, USA

Unsupervised speech disentanglement aims at separating fast varying from slowly varying components of a speech signal. In this contribution, we take a closer look at the embedding vector representing the slowly varying signal components, commonly named the speaker embedding vector. We ask, which properties of a speaker's voice are captured and investigate to which extent do individual embedding vector components sign responsible for them, using the concept of Shapley values. Our findings show that certain speaker-specific acoustic-phonetic properties can be fairly well predicted from the speaker embedding, while the investigated more abstract voice quality features cannot., Comment: Presented at the ITG conference on Speech Communication 2023

Confidentiality, integrity protection, and high availability, abbreviated to CIA, are essential properties for trustworthy data systems. The rise of cloud computing and the growing demand for multiparty applications however means that building modern CIA systems is more challenging than ever. In response, we present the Confidential Consortium Framework (CCF), a general-purpose foundation for developing secure stateful CIA applications. CCF combines centralized compute with decentralized trust, supporting deployment on untrusted cloud infrastructure and transparent governance by mutually untrusted parties. CCF leverages hardware-based trusted execution environments for remotely verifiable confidentiality and code integrity. This is coupled with state machine replication backed by an auditable immutable ledger for data integrity and high availability. CCF enables each service to bring its own application logic, custom multiparty governance model, and deployment scenario, decoupling the operators of nodes from the consortium that governs them. CCF is open-source and available now at https://github.com/microsoft/CCF., Comment: 16 pages, 9 figures. To appear in the Proceedings of the VLDB Endowment, Volume 17

We study the Cahn-Hilliard-Biot model with respect to its mathematical well-posedness. The system models flow through deformable porous media in which the solid material has two phases with distinct material properties. The two phases of the porous material evolve according to a generalized Ginzburg-Landau energy functional, with additional influence from both viscoelastic and fluid effects. The flow-deformation coupling in the system is governed by Biot's theory. This results in a three-way coupled system that can be viewed as an extension of the Cahn-Larche equations by adding a fluid flowing through the medium. We distinguish the cases between a spatially dependent and a state-dependent Biot-Willis function. In the latter case, we consider a regularized system. In both cases, we use a Galerkin approximation to discretize the system and derive suitable energy estimates. Moreover, we apply compactness methods to pass to the limit in the discretized system. In the case of Vegard's law and homogeneous elasticity, we show that the weak solution depends continuously on the data and is unique. Lastly, we present some numerical simulations to highlight the features of the system as a tumor growth model.

One of the hallmark properties of fluids is their shear viscosity which is, among other things, responsible for parabolic flow profiles through narrow channels. In recent years, there has been a growing number of observations of said flow profiles in electronic transport measurements in a variety of material systems, most notably in graphene. In this paper, we investigate the shear viscosity of interacting graphene from a theoretical point of view. We study both a phenomenological as well as a microscopic model and find excellent agreement between the two. Our main finding is collective modes make a sizeable contribution to the viscosity that can equal or even outweigh the electronic contribution that is usually assumed dominant. We comment on how this finding carries over to systems beyond graphene and related Dirac materials.

West Nile virus (WNV) is prevalent in the United States but it shows considerable divergence in transmission patterns and spatio-temporal intensity.It is to be noted that the mechanism that drives the transmission potential of WNV is described by the abilities of host species to maintain and disseminate the pathogens pertinent with different eco-epidemiological factors that have an influence on the contact rates amongst the interacting species.There is growing evidence that several vectors exhibit strong feeding preferences towards different host communities.We construct a process based weather driven ordinary differential equation (ODE) model to understand the impact of one vector species Culex pipiens, preferred avian and non-preferred human hosts and compared it surveillance data for the Culex pipiens complex collected in Cook County, Illinois, USA.In our mechanistic model, we also demonstrate that adulticide treatments produced significant reductions in the Culex pipiens population.We take into account the feeding index that can be described as the ratio between observed frequency of mosquitoes feeding on one host compared to another host, divided by the expected frequency of mosquitoes feeding on these two hosts based on the presence of the particular hosts to develop this transmission model for WNV. Our findings demonstrate that the interplay between the feeding index and mosquito abatement strategy is rather a complex phenomenon and it induces a heterogeneous contact rates that should be included while modelling multi-host, multi-vector transmission model.

Anomaly Detection (AD) is a critical task that involves identifying observations that do not conform to a learned model of normality. Prior work in deep AD is predominantly based on a familiarity hypothesis, where familiar features serve as the reference in a pre-trained embedding space. While this strategy has proven highly successful, it turns out that it causes consistent false negatives when anomalies consist of truly novel features that are not well captured by the pre-trained encoding. We propose a novel approach to AD using explainability to capture such novel features as unexplained observations in the input space. We achieve strong performance across a wide range of anomaly benchmarks by combining familiarity and novelty in a hybrid approach. Our approach establishes a new state-of-the-art across multiple benchmarks, handling diverse anomaly types while eliminating the need for expensive background models and dense matching. In particular, we show that by taking account of novel features, we reduce false negative anomalies by up to 40% on challenging benchmarks compared to the state-of-the-art. Our method gives visually inspectable explanations for pixel-level anomalies.

There is an growing interest in using Large Language Models (LLMs) in multi-agent systems to tackle interactive real-world tasks that require effective collaboration and assessing complex situations. Yet, we still have a limited understanding of LLMs' communication and decision-making abilities in multi-agent setups. The fundamental task of negotiation spans many key features of communication, such as cooperation, competition, and manipulation potentials. Thus, we propose using scorable negotiation to evaluate LLMs. We create a testbed of complex multi-agent, multi-issue, and semantically rich negotiation games. To reach an agreement, agents must have strong arithmetic, inference, exploration, and planning capabilities while integrating them in a dynamic and multi-turn setup. We propose multiple metrics to rigorously quantify agents' performance and alignment with the assigned role. We provide procedures to create new games and increase games' difficulty to have an evolving benchmark. Importantly, we evaluate critical safety aspects such as the interaction dynamics between agents influenced by greedy and adversarial players. Our benchmark is highly challenging; GPT-3.5 and small models mostly fail, and GPT-4 and SoTA large models (e.g., Llama-3 70b) still underperform., Comment: Updated version with major additions (new experiments, evaluation, and attacks)

Machine learning has been successfully applied to grid-based PDE modeling in various scientific applications. However, learned PDE solvers based on Lagrangian particle discretizations, which are the preferred approach to problems with free surfaces or complex physics, remain largely unexplored. We present LagrangeBench, the first benchmarking suite for Lagrangian particle problems, focusing on temporal coarse-graining. In particular, our contribution is: (a) seven new fluid mechanics datasets (four in 2D and three in 3D) generated with the Smoothed Particle Hydrodynamics (SPH) method including the Taylor-Green vortex, lid-driven cavity, reverse Poiseuille flow, and dam break, each of which includes different physics like solid wall interactions or free surface, (b) efficient JAX-based API with various recent training strategies and three neighbor search routines, and (c) JAX implementation of established Graph Neural Networks (GNNs) like GNS and SEGNN with baseline results. Finally, to measure the performance of learned surrogates we go beyond established position errors and introduce physical metrics like kinetic energy MSE and Sinkhorn distance for the particle distribution. Our codebase is available at https://github.com/tumaer/lagrangebench ., Comment: Accepted at 37th Conference on Neural Information Processing Systems (NeurIPS 2023) Track on Datasets and Benchmarks

The availability of rich and vast data sources has greatly advanced machine learning applications in various domains. However, data with privacy concerns comes with stringent regulations that frequently prohibited data access and data sharing. Overcoming these obstacles in compliance with privacy considerations is key for technological progress in many real-world application scenarios that involve privacy sensitive data. Differentially private (DP) data publishing provides a compelling solution, where only a sanitized form of the data is publicly released, enabling privacy-preserving downstream analysis and reproducible research in sensitive domains. In recent years, various approaches have been proposed for achieving privacy-preserving high-dimensional data generation by private training on top of deep neural networks. In this paper, we present a novel unified view that systematizes these approaches. Our view provides a joint design space for systematically deriving methods that cater to different use cases. We then discuss the strengths, limitations, and inherent correlations between different approaches, aiming to shed light on crucial aspects and inspire future research. We conclude by presenting potential paths forward for the field of DP data generation, with the aim of steering the community toward making the next important steps in advancing privacy-preserving learning.

Galaxy clusters, which underwent a recent ($\leq3$ Gyr) major merger, offer a harsher environment due to the global hydrodynamical disturbance and the merger-shock heated ICM. However, the aftermath of such extreme cluster interactions on the member galaxy properties is not very well constrained. We explore the integrated star formation properties of galaxies through galaxy colours, as well as morphology buildup in three nearby ($0.04

IceCube alert events are neutrinos with a moderate-to-high probability of having astrophysical origin. In this study, we analyze 11 years of IceCube data and investigate 122 alert events and a selection of high-energy tracks detected between 2009 and the end of 2021. This high-energy event selection (alert events + high-energy tracks) has an average probability of $\geq 0.5$ to be of astrophysical origin. We search for additional continuous and transient neutrino emission within the high-energy events' error regions. We find no evidence for significant continuous neutrino emission from any of the alert event directions. The only locally significant neutrino emission is the transient emission associated with the blazar TXS~0506+056, with a local significance of $ 3 \sigma$, which confirms previous IceCube studies. When correcting for 122 test positions, the global p-value is $0.156$ and is compatible with the background hypothesis. We constrain the total continuous flux emitted from all 122 test positions at 100~TeV to be below $1.2 \times 10^{-15}$~(TeV cm$^2$ s)$^{-1}$ at 90% confidence assuming an $E^{-2}$ spectrum. This corresponds to 4.5% of IceCube's astrophysical diffuse flux. Overall, we find no indication that alert events, in general, are linked to lower-energetic continuous or transient neutrino emission., Comment: Accepted by ApJ

K isomers in transuranium nuclei have become a most interesting subject in nuclear structure investigation in many laboratories within the recent twenty years. In this paper an overview on the present day situation will be given. It will focus on the conditions for occuring of this kind of isomers, their decay properties and systematics in their properties as far as experimental data are available.

Despite recent success, state-of-the-art learning-based models remain highly vulnerable to input changes such as adversarial examples. In order to obtain certifiable robustness against such perturbations, recent work considers Lipschitz-based regularizers or constraints while at the same time increasing prediction margin. Unfortunately, this comes at the cost of significantly decreased accuracy. In this paper, we propose a Calibrated Lipschitz-Margin Loss (CLL) that addresses this issue and improves certified robustness by tackling two problems: Firstly, commonly used margin losses do not adjust the penalties to the shrinking output distribution; caused by minimizing the Lipschitz constant $K$. Secondly, and most importantly, we observe that minimization of $K$ can lead to overly smooth decision functions. This limits the model's complexity and thus reduces accuracy. Our CLL addresses these issues by explicitly calibrating the loss w.r.t. margin and Lipschitz constant, thereby establishing full control over slack and improving robustness certificates even with larger Lipschitz constants. On CIFAR-10, CIFAR-100 and Tiny-ImageNet, our models consistently outperform losses that leave the constant unattended. On CIFAR-100 and Tiny-ImageNet, CLL improves upon state-of-the-art deterministic $L_2$ robust accuracies. In contrast to current trends, we unlock potential of much smaller models without $K=1$ constraints., Comment: To be published at GCPR 2023

The transmission of strong laser light in nonlinear optical materials can generate output photons sources that carry quantum entanglement in multiple degrees of freedom, making this process a fundamentally important tool in optical quantum technology. However, the availability of efficient optical crystals for entangled light generation is severely limited in terms of diversity, thus reducing the prospects for the implementation of next-generation protocols in quantum sensing, communication and computing. To overcome this, we developed and implemented a multi-scale first-principles modeling technique for the computational discovery of novel nonlinear optical devices based on metal-organic framework (MOF) materials that can efficiently generate entangled light via spontaneous parametric down-conversion(SPDC). Using collinear degenerate type-I SPDC as a case study, we computationally screen a database of 114,373 synthesized MOF materials to establish correlations between the structure and chemical composition of MOFs with the brightness and coherence properties of entangled photon pairs. We identify a subset of 49 non-centrosymmetric mono-ligand MOF crystals with high chemical and optical stability that produce entangled photon pairs with intrinsic $G^{(2)}$ correlation times $\tau_c\sim 10-30$ fs and pair generation rates in the range $10^4-10^{8}$ s$^{-1}$mW$^{-1}$mm$^{-1}$ at 1064 nm. Conditions for optimal type-I phase matching are given for each MOF and relationships between pair brightness, crystal band gap and optical birefringence are discussed. Correlations between the optical properties of crystals and their constituent molecular ligands are also given. Our work paves the way for the computational design of MOF-based devices for optical quantum technology., Comment: Supplementary Material (16 pages, 14 figures, 2 table) at the end of manuscript. Github (https://github.com/snoozynooj/Type-I-SPDC)

We investigate several panchromatic scaling relations (SRs) for the dwarf irregular galaxy NGC 1569 using IFU data from the Metal-THINGS Survey. Among the spatially resolved properties analyzed, we explore SRs between the stellar mass, SFR, molecular gas, total gas, baryonic mass, gas metallicity, gas fraction, SFE and effective oxygen yields. Such multiwavelength SRs are analyzed at a spatial resolution of 180 pc, by combining our IFU observations with data from the surveys THINGS, CARMA, and archival data from DustPedia. Although we recover several known relations, our slopes are different to previously reported ones. Our star formation main sequence, Kennicutt-Schmidt (KS) and molecular KS relations show higher SFRs, lower scatter, and higher correlations, with steeper (1.21), and flatter slopes (0.96, 0.58) respectively. The shape of the SRs including metallicity, stellar mass, and gas fraction are flat, with an average value of 12+log(O/H) $\sim$ 8.12 dex. The baryonic mass vs effective oxygen yields, and the stellar, gas and baryonic mass vs SFE show higher dispersions and lower correlations. Since we use the dust mass as a tracer of gas mass, we derive the Dust-to-Gas Ratio and the CO luminosity-to-molecular gas mass conversion factors, showing differences of 0.16 and 0.95 dex for the total and molecular gas surface density, respectively, in comparison to previously reported values. We use a self regulated feedback model to conclude that stellar feedback plays an important role generating outflows in NGC 1569.

We report on the serendipitous discovery of an unprecedented interaction between the radio lobe of a radio galaxy and a spiral galaxy. The discovery was made thanks to LOFAR observations at 144 MHz of the galaxy cluster Abell 160 ($z=0.04317$) provided by the LOFAR Two-metre Sky Survey. The new low-frequency observations revealed that one of the radio plumes of the central galaxy GIN 049 overlaps with the spiral galaxy JO36. Previous studies carried out with MUSE revealed that the warm ionized gas in the disk of JO36, traced by the H$\alpha$ emission, is severely truncated with respect to the stellar disk. We further explore this unique system by including new uGMRT observations at 675 MHz to map the spectral index. The emerging scenario is that JO36 has interacted with the radio plume in the past 200-500 Myr. The encounter resulted in a positive feedback event for JO36 in the form of a star formation rate burst of $\sim14$ $M_\odot$ yr$^{-1}$. In turn, the galaxy passage left a trace in the radio-old plasma by re-shaping the old relativistic plasma via magnetic draping., Comment: 20 pages, 11 figures. Accepted for publication on ApJ on September 4th, 2023

Cluster galaxies are affected by the surrounding environment, which influences, in particular, their gas, stellar content and morphology. In particular, the ram-pressure exerted by the intracluster medium promotes the formation of multi-phase tails of stripped gas detectable both at optical wavelengths and in the sub-mm and radio regimes, tracing the cold molecular and atomic gas components, respectively. In this work we analyze a sample of sixteen galaxies belonging to clusters at redshift $\sim 0.05$ showing evidence of an asymmetric HI morphology (based on MeerKAT observations) with and without a star forming tail. To this sample we add three galaxies with evidence of a star forming tail and no HI detection. Here we present the galaxies $\rm H_{2}$ gas content from APEX observations of the CO(2-1) emission. We find that in most galaxies with a star forming tail the $\rm H_{2}$ global content is enhanced with respect to undisturbed field galaxies with similar stellar masses, suggesting an evolutionary path driven by the ram-pressure stripping. As galaxies enter into the clusters their HI is displaced but also partially converted into $\rm H_{2}$, so that they are $\rm H_{2}$ enriched when they pass close to the pericenter, i. e. when they develop the star forming tails that are visible in UV/B broad bands and in H$\alpha$ emission. An inspection of the phase-space diagram for our sample suggests an anticorrelation between the HI and $\rm H_{2}$ gas phases as galaxies fall into the cluster potential. This peculiar behaviour is a key signature of the ram-pressure stripping in action., Comment: 19 pages, 8 figures, accepted for publication in ApJ

We investigate how the ageing of stellar populations can drive a morphological transformation in galaxies whose star formation (SF) activity has been quenched on short timescales, like in cluster galaxies subject to ram pressure stripping from the intracluster medium. For this purpose, we use a sample of 91 galaxies with MUSE data from the GASP program and of their spatially resolved SF history derived with the spectral modelling software SINOPSIS. We simulate the future continuation of the SF activities by exploring two quenching scenarios: an instantaneous truncation of the SF across the whole disc, and an outside-in quenching with typical stripping timescales of 0.5 Gyr and 1 Gyr. For each scenario we produce mock MUSE spectroscopic datacubes and optical images for our galaxies during their evolution, and classify their morphology using a new diagnostic tool, calibrated on cluster galaxies from the OmegaWINGS Survey. We find that, in all scenarios considered, the initial galaxy population dominated by blue-cloud spirals (90%) evolves into a mixed population mostly composed by red-sequence spirals (50-55%) and lenticulars (~40%). The morphology transformation is completed after just 1.5-3.5 Gyr, proceeding faster in more efficient quenching scenarios. Our results indicate that, even without accounting for dynamical processes, SF quenching caused by the harsh environment of a cluster can significantly affect the morphology of the infalling galaxy population on timescales of a few Gyr., Comment: 19 pages, 11 figures, accepted for publication in MNRAS