arXiv subjects

Stefano Berti

Publications and source records attributed to Stefano Berti.

At least 19 recordsLinked to original sources

arXiv2026

Vertical transport and confinement of weakly buoyant particles in a convective ocean mixed-layer model

Understanding how particles such as microplastics and plankton spread, or accumulate in localized regions below the ocean surface is crucial both for the development of sea pollutant control strategies and for marine ecology. Yet the role of vertically inhomogeneous turbulence, due to the vertical structure of density stratification in the ocean, remains poorly understood. We study, by means of direct numerical simulations, the dynamics of weakly inertial, quasi-neutrally buoyant particles in an idealized two-layer fluid model aimed to represent the ocean convective mixed layer and the more stably stratified upper thermocline beneath. We find that particles eventually rise at the surface if their density matches the mean value of the fluid at a selected depth within the mixed layer. Interestingly, however, for all other values of density, they accumulate around their neutral-buoyancy depth within the thermocline, where stratification hinders vertical transport. Our analysis shows that the thickness of the accumulation layer results from the competition between buoyancy-driven confinement, internal-wave motions, and turbulent fluctuations penetrating from the mixed layer. Smaller particles, which respond more rapidly to flow fluctuations, form broader layers; layers also broaden as their equilibrium depth approaches the mixed layer. We then derive a reduced stochastic model that explains the dependence on particle inertia and reference depth of such spreading around the equilibrium position. These results show how the coupled action of stratification and decaying mixed-layer turbulence controls subsurface particle-layer formation, which may be relevant to ocean-biology oriented studies and to improve the sampling of plastic pollution in the sea.

physics.flu-dyn↗

arXiv2026

Temporal Tactile Encoding and Compliance for Intent-Aware Robot-to-Human Bimanual Handover

Reliable robot-to-human handover requires the robot to infer when the person is ready to receive the object, and release it safely, comfortably, and at the right time. This is challenging because visual observations alone may not disambiguate clear taking intent from accidental contact, weak grasping, wrong-direction forces, or transient interactions. In this work we treat human-robot handover as an intrinsically multimodal problem. Our approach couples a VLA model with a compliance controller that reduces interaction forces during object transfer. We finetune the VLA model with human demonstrations using RGB observation, temporally encoded tactile feedback and proprioception. We evaluate the complete system in a human-subject study against two baselines: one without tactile feedback and one using tactile feedback without compliance control. We hypothesize that combining compliance and temporal tactile encoding yields the most reliable and comfortable handovers, as compliance facilitates physical interaction while tactile history captures sustained taking intent. Performance is measured through objective metrics and an ad-hoc questionnaire. The results show that the two components provide complementary benefits and substantially outperform the baselines. Code and data will be released upon acceptance.

cs.RO↗

arXiv2026

Nonhomogeneous elastic turbulence in the two-dimensional Taylor-Couette flow

Elastic turbulence is a spatially and temporally disordered flow state appearing in viscoelastic fluids at vanishing fluid inertia and large elasticity. The resulting flows have broad technological interest, particularly to enhance mixing and heat transfer in microdevices. Although its experimental characterization is now well established in different setups, its theoretical understanding and numerical reproducibility remain challenging, especially in wall-bounded geometries. By means of numerical simulations, we investigate the onset of elastic turbulence and the characteristics of the developed turbulent-like states in the two-dimensional, confined, Taylor-Couette system. First, we characterize the purely elastic instability, addressing previously contrasting evidences. We then show that the fully nonlinear dynamics are weakly anisotropic and strongly nonhomogeneous. Indeed, they are confined in a dynamically active region adjacent to the inner wall, akin to the elastic boundary layer from previous predictions. Within this region,in spite of some non-negligible deviations due to the nonhomeogeneity of our setup, the statistical and spectral turbulent properties are to reasonable extent not far from the theoretical expectations and experimental observations.

physics.flu-dyn↗

arXiv2026

A Baseline Study and Benchmark for Few-Shot Open-Set Action Recognition with Feature Residual Discrimination

Few-Shot Action Recognition (FS-AR) has shown promising results but is often limited by a closed-set assumption that fails in real-world open-set scenarios. While Few-Shot Open-Set (FSOS) recognition is well-established for images, its extension to spatio-temporal video data remains underexplored. To address this, we propose an architectural extension based on a Feature-Residual Discriminator (FR-Disc), adapting previous work on skeletal data to the more complex video domain. Extensive experiments on five datasets demonstrate that while common open-set techniques provide only marginal gains, our FR-Disc significantly enhances unknown rejection capabilities without compromising closed-set accuracy, setting a new state-of-the-art for FSOS-AR. The project website, code, and benchmark are available at: https://hsp-iit.github.io/fsosar/.

cs.CV↗

arXiv2025

Effects of high-frequency and balanced motions on Lagrangian pair dispersion at the ocean surface

We investigate the properties of relative dispersion of Lagrangian particles in a global-ocean simulation resolving both inertia-gravity waves (IGW) and meso and submesoscale (M/SM) turbulence. More specifically, we test if the dispersion laws depend on the shape of the Eulerian kinetic energy spectrum, as predicted from quasi-geostrophic turbulence theory. To this end, we focus on two areas, in the Kuroshio Extension and in the Gulf Stream, for which the relative importance of IGW compared to M/SM vary in summer and winter. In winter, Lagrangian statistical indicators return a picture in overall agreement with the shape of the kinetic energy spectrum. Conversely, in summer, when submesoscales are less energetic and higher-frequency internal waves gain importance, the expected relations between dispersion properties and spectra do not seem to hold. This apparent discrepancy is explained by decomposing the flow into nearly-balanced motions and internal gravity waves, and showing that the latter dominate the kinetic energy spectrum at small scales. Our results are consistent with the hypothesis that high-frequency IGWs do not impact relative dispersion, which is then controlled by the nearly-balanced, mainly rotational, flow component at larger scales. These results highlight that geostrophic velocities derived from wide-swath altimeters, such as SWOT, may present limits when estimating surface dispersion, and that current measuring satellite missions may provide the complementary information to do so.

physics.flu-dyn↗

arXiv2024

Impact of ageostrophic dynamics on the predictability of Lagrangian trajectories in surface-ocean turbulence

Turbulent flows at the surface of the ocean deviate from geostrophic equilibrium on scales smaller than about 10 km. These scales are associated with important vertical transport of active and passive tracers, and should play a prominent role in the heat transport at climatic scales and for plankton dynamics. Measuring velocity fields on such small scales is notoriously difficult but new, high-resolution satellite altimetry is starting to reveal them. However, the satellite-derived velocities essentially represent the geostrophic flow component, and the impact of unresolved ageostrophic motions on particle dispersion needs to be understood to properly characterize transport properties. Here, we investigate ocean fine-scale turbulence using a model that represents some of the processes due to ageostrophic dynamics. We take a Lagrangian approach and focus on the predictability of the particle dynamics, comparing trajectories advected by either the full flow or by its geostrophic component only. Our results indicate that, over long times, relative dispersion is marginally affected by the filtering of the ageostrophic component. Nevertheless, advection by the filtered flow leads to an overestimation of the typical pair-separation rate, and to a bias on trajectories (in terms of displacement from the actual ones), whose importance grows with the Rossby number. We further explore the intensity of the transient particle clustering induced by ageostrophic motions and find that it can be significant, even for small flow compressibility. Indeed, we show that clustering is here due to the interplay between compressibility and persistent flow structures that trap particles, enhancing their aggregation.

physics.flu-dyn↗

arXiv2024

The impact of Compositionality in Zero-shot Multi-label action recognition for Object-based tasks

Addressing multi-label action recognition in videos represents a significant challenge for robotic applications in dynamic environments, especially when the robot is required to cooperate with humans in tasks that involve objects. Existing methods still struggle to recognize unseen actions or require extensive training data. To overcome these problems, we propose Dual-VCLIP, a unified approach for zero-shot multi-label action recognition. Dual-VCLIP enhances VCLIP, a zero-shot action recognition method, with the DualCoOp method for multi-label image classification. The strength of our method is that at training time it only learns two prompts, and it is therefore much simpler than other methods. We validate our method on the Charades dataset that includes a majority of object-based actions, demonstrating that -- despite its simplicity -- our method performs favorably with respect to existing methods on the complete dataset, and promising performance when tested on unseen actions. Our contribution emphasizes the impact of verb-object class-splits during robots' training for new cooperative tasks, highlighting the influence on the performance and giving insights into mitigating biases.

cs.CV↗

arXiv2023

Particle dispersion and clustering in surface ocean turbulence with ageostrophic dynamics

Upper-ocean turbulent flows at horizontal length scales smaller than the deformation radius depart from geostrophic equilibrium and develop important vertical velocities, which are key to marine ecology and climatic processes. Due to their small size and fast temporal evolution, these fine scales are difficult to measure during oceanographic campaigns. Instruments such as Lagrangian drifters have provided another way to characterize these scales through the analysis of pair-dispersion evolution, and have pointed out striking particle convergence events. By means of numerical simulations, we investigate such processes in a model of surface-ocean turbulence that includes ageostrophic motions. This model originates from a Rossby-number expansion of the primitive equations and reduces to the surface quasi-geostrophic model, a paradigm of submesoscale dynamics, in the limit of vanishing Rossby number. We focus on the effect of the ageostrophic dynamics on the pair-dispersion and clustering properties of Lagrangian tracer particles at the ocean surface. Our results indicate that while over long times the pair separation process is barely affected by the ageostrophic component of the velocity field, the latter is responsible for the formation of temporary particle aggregates, and the intensity of this phenomenon increases with the Rossby number. We further show that Lagrangian tracers preferentially accumulate in cyclonic frontal regions, which is in agreement with observations and other more realistic modeling studies. These findings appear interesting to improve the understanding of the turbulent transport by ocean fine scales, and in light of upcoming, new high-resolution satellite data of surface velocity fields.

physics.flu-dyn↗

arXiv2022

Surface light modulation by sea ice and phytoplankton survival in a convective flow model

Plankton dynamics depend in a complex manner on a variety of physical phenomena, according to both experimental and numerical data. In particular, experimental field studies have highlighted the relation between phytoplankton survival and turbulent upwelling and downwelling from thermal convection. Recent numerical works have also shown the importance of accounting for advective transport by persistent structures in simulation models. In nutrient-rich polar marine environments phytoplankton blooms are critically limited by light availability under ice-covered waters. Such heterogeneity of the light intensity distribution, in association with a large-scale coherent fluid flow, can give rise to nontrivial growth dynamics. In this work we extend a previous advection-reaction-diffusion model of phytoplankton light-limited vertical dynamics in the presence of convective transport. Specifically, we consider horizontally heterogeneous light conditions through the use of two regions with different production regimes, modelling the absence (presence) of light under (in between) obstacles. Such a model is intended as an idealized representation of nonuniformly ice-covered polar waters. By means of numerical simulations, we find that the main role of advective transport is to hinder phytoplankton growth, but also that such effect depends on the positions of the obstacles with respect to the upwelling and downwelling flow regions. Furthermore, we show that the sinking speed due to the density difference between phytoplankton organisms and water, while small, plays an important role, which depends on how it adds to the flow. These results indicate that advective transport can have a crucial impact on the survival conditions of sinking phytoplankton species in polar environments.

physics.flu-dyn↗

arXiv2022

Towards Confidence-guided Shape Completion for Robotic Applications

Many robotic tasks involving some form of 3D visual perception greatly benefit from a complete knowledge of the working environment. However, robots often have to tackle unstructured environments and their onboard visual sensors can only provide incomplete information due to limited workspaces, clutter or object self-occlusion. In recent years, deep learning architectures for shape completion have begun taking traction as effective means of inferring a complete 3D object representation from partial visual data. Nevertheless, most of the existing state-of-the-art approaches provide a fixed output resolution in the form of voxel grids, strictly related to the size of the neural network output stage. While this is enough for some tasks, e.g. obstacle avoidance in navigation, grasping and manipulation require finer resolutions and simply scaling up the neural network outputs is computationally expensive. In this paper, we address this limitation by proposing an object shape completion method based on an implicit 3D representation providing a confidence value for each reconstructed point. As a second contribution, we propose a gradient-based method for efficiently sampling such implicit function at an arbitrary resolution, tunable at inference time. We experimentally validate our approach by comparing reconstructed shapes with ground truths, and by deploying our shape completion algorithm in a robotic grasping pipeline. In both cases, we compare results with a state-of-the-art shape completion approach.

cs.CV↗

arXiv2022

One-Shot Open-Set Skeleton-Based Action Recognition

Action recognition is a fundamental capability for humanoid robots to interact and cooperate with humans. This application requires the action recognition system to be designed so that new actions can be easily added, while unknown actions are identified and ignored. In recent years, deep-learning approaches represented the principal solution to the Action Recognition problem. However, most models often require a large dataset of manually-labeled samples. In this work we target One-Shot deep-learning models, because they can deal with just a single instance for class. Unfortunately, One-Shot models assume that, at inference time, the action to recognize falls into the support set and they fail when the action lies outside the support set. Few-Shot Open-Set Recognition (FSOSR) solutions attempt to address that flaw, but current solutions consider only static images and not sequences of images. Static images remain insufficient to discriminate actions such as sitting-down and standing-up. In this paper we propose a novel model that addresses the FSOSR problem with a One-Shot model that is augmented with a discriminator that rejects unknown actions. This model is useful for applications in humanoid robotics, because it allows to easily add new classes and determine whether an input sequence is among the ones that are known to the system. We show how to train the whole model in an end-to-end fashion and we perform quantitative and qualitative analyses. Finally, we provide real-world examples.

cs.RO↗

arXiv2021

Three-dimensional turbulence effects on plankton dynamics behind an obstacle

We study a predator-prey model of plankton dynamics in the two and three-dimensional wakes of turbulent flows behind a cylinder, focusing on the impact of the three-dimensional character of the carrying velocity field on population variance spectra and spatial distributions. By means of direct numerical simulations, we find that the qualitative behavior of the biological dynamics is mostly independent of the space dimensionality, which suggests that only the relation between the typical flow and biological timescales is crucial to observe persistent blooms. Similarly, in both cases, we find that the spectral properties of the planktonic populations are essentially indistinguishable from those of an inert tracer. The main difference arising from the comparison of the two and three-dimensional configurations concerns the local spatial distribution of plankton density fields. Indeed, the three-dimensional turbulent dynamics tend to destroy the localized coherent structures characterizing the two-dimensional flow, in which the planktonic species are mostly concentrated, thus reducing the phytoplankton biomass in the system.

physics.flu-dyn↗

arXiv2021

Effects of large-scale advection and small-scale turbulent diffusion on vertical phytoplankton dynamics

Turbulence has been recognized as a factor of paramount importance for the survival or extinction of sinking phytoplankton species. However, dealing with its multiscale nature in models of coupled fluid and biological dynamics is a formidable challenge. Advection by coherent structures, as those related to winter convection and Langmuir circulation, is also recognized to play a role in the survival and localization of phytoplankton. In this work we revisit a theoretically appealing model for phytoplankton vertical dynamics, and numerically investigate how large-scale fluid motions affect the survival conditions and the spatial distribution of the biological population. For this purpose, and to work with realistic parameter values, we adopt a kinematic flow field to account for the different spatial and temporal scales of turbulent motions. The dynamics of the population density are described by an advection-reaction-diffusion model with a spatially heterogeneous growth term proportional to sunlight availability. We explore the role of fluid transport by progressively increasing the complexity of the flow in terms of spatial and temporal scales. We find that, due to the large-scale circulation, phytoplankton accumulates in downwelling regions and its growth is reduced, confirming previous indications in slightly different conditions. We then explain the observed phenomenology in terms of a plankton filament model. Moreover, by contrasting the results in our different flow cases, we show that the large-scale coherent structures have an overwhelming importance. Indeed, we find that smaller-scale motions only quite weakly affect the dynamics, without altering the general mechanism identified. Such results are relevant for parameterizations in numerical models of phytoplankton life cycles in realistic oceanic flow conditions.

physics.flu-dyn↗

arXiv2021

Statistical properties of two-dimensional elastic turbulence

We numerically investigate the spatial and temporal statistical properties of a dilute polymer solution in the elastic turbulence regime, i.e., in the chaotic flow state occurring at vanishing Reynolds and high Weissenberg numbers. We aim at elucidating the relations between measurements of flow properties performed in the spatial domain with the ones taken in the temporal domain, which is a key point for the interpretation of experimental results on elastic turbulence and to discuss the validity of Taylor's hypothesis. To this end, we carry out extensive direct numerical simulations of the two-dimensional Kolmogorov flow of an Oldroyd-B viscoelastic fluid. Static point-like numerical probes are placed at different locations in the flow, particularly at the extrema of mean flow amplitude. The results in the fully developed elastic turbulence regime reveal large velocity fluctuations, as compared to the mean flow, leading to a partial breakdown of Taylor's frozen-field hypothesis. While second-order statistics, probed by spectra and structure functions, display consistent scaling behaviors in the spatial and temporal domains, the third-order statistics highlight robust differences. In particular the temporal analysis fails to capture the skewness of streamwise longitudinal velocity increments. Finally, we assess both the degree of statistical inhomogeneity and isotropy of the flow turbulent fluctuations as a function of scale. While the system is only weakly non-homogenous in the cross-stream direction, it is found to be highly anisotropic at all scales.

physics.flu-dyn↗

arXiv2021

Calliope -- A Polyphonic Music Transformer

The polyphonic nature of music makes the application of deep learning to music modelling a challenging task. On the other hand, the Transformer architecture seems to be a good fit for this kind of data. In this work, we present Calliope, a novel autoencoder model based on Transformers for the efficient modelling of multi-track sequences of polyphonic music. The experiments show that our model is able to improve the state of the art on musical sequence reconstruction and generation, with remarkably good results especially on long sequences.

cs.SD↗

arXiv2021

Predator-prey plankton dynamics in turbulent wakes behind islands

Plankton constitutes the productive base of aquatic ecosystems and plays an essential role in the global carbon cycle. The impact of hydrodynamic conditions on the biological activity of plankton species can manifest in a variety of different ways and the understanding of the basic effects due to multiscale complex flows still appears incomplete. Here, we consider a predator-prey model of plankton dynamics in the presence of a turbulent flow past an idealized island, to investigate the conditions under which an algal bloom is observed, and the resulting patchiness of plankton distributions. By means of direct numerical simulations, we explore the role of the turbulent intensity and of the island shape. In particular, we focus on population variance spectra, and on their relation with the statistical features of the turbulent flow, as well as on the correlation between the spatial distribution of the planktonic species and velocity field persistent structures. We find that both the average biomass and local plankton dynamics critically depend on the relation between advective and biological time scales, confirming and extending previous results obtained in simpler flow settings. We further provide evidence that, in the present system, due to local flow action plankton accumulates in localized filamentary regions. Small turbulent scales are found to impact the statistics of plankton density fields at very fine scales, but the main global features of the population dynamics only weakly depend on the Reynolds number and are also found to be remarkably independent of the geometrical details of the obstacle.

physics.flu-dyn↗

arXiv2021

Lagrangian pair dispersion in upper-ocean turbulence in the presence of mixed-layer instabilities

Turbulence in the upper ocean in the submesoscale range (scales smaller than the deformation radius) plays an important role for the heat exchange with the atmosphere and for oceanic biogeochemistry. Its dynamics should strongly depend on the seasonal cycle and the associated mixed-layer instabilities. The latter are particularly relevant in winter and are responsible for the formation of energetic small scales that extend over the whole depth of the mixed layer. The knowledge of the transport properties of oceanic flows at depth, which is essential to understand the coupling between surface and interior dynamics, however, is still limited. By means of numerical simulations, we explore the Lagrangian dispersion properties of turbulent flows in a quasi-geostrophic model system allowing for both thermocline and mixed-layer instabilities. The results indicate that, when mixed-layer instabilities are present, the dispersion regime is local from the surface down to depths comparable with that of the interface with the thermocline, while in their absence dispersion quickly becomes nonlocal versus depth. We then identify the origin of such behavior in the existence of fine-scale energetic structures due to mixed-layer instabilities. We further discuss the effect of vertical shear on the Lagrangian particle spreading and address the correlation between the dispersion properties at the surface and at depth, which is relevant to assess the possibility of inferring the dynamical features of deeper flows from the more accessible surface ones.

physics.flu-dyn↗

arXiv2021

Elastic turbulence in two-dimensional cross-slot viscoelastic flows

We report evidence of irregular unsteady flow of two-dimensional polymer solutions in the absence of inertia in cross-slot geometry using numerical simulations of Oldroyd-B model. By exploring the transition to time-dependent flow versus both the fluid elasticity and the polymer concentration, we find periodic behaviour close to the instability threshold and more complex flows at larger elasticity, in agreement with experimental findings. For high enough elasticity we obtain dynamics pointing to elastic turbulence, with temporal spectra of velocity fluctuations showing a power-law decay, of exponent in between -3 and -2, and probability density functions of velocity fluctuations that weakly deviate from Gaussianity while high non-Gaussian tails characterise those of local accelerations.

cond-mat.soft↗