Search

Referring Video Object Segmentation (RVOS) relies on natural language expressions to segment an object in a video clip. Existing methods restrict reasoning either to independent short clips, losing global context, or process the entire video offline, impairing their application in a streaming fashion. In this work, we aim to surpass these limitations and design an RVOS method capable of effectively operating in streaming-like scenarios while retaining contextual information from past frames. We build upon the Segment-Anything 2 (SAM2) model, that provides robust segmentation and tracking capabilities and is naturally suited for streaming processing. We make SAM2 wiser, by empowering it with natural language understanding and explicit temporal modeling at the feature extraction stage, without fine-tuning its weights, and without outsourcing modality interaction to external models. To this end, we introduce a novel adapter module that injects temporal information and multi-modal cues in the feature extraction process. We further reveal the phenomenon of tracking bias in SAM2 and propose a learnable module to adjust its tracking focus when the current frame features suggest a new object more aligned with the caption. Our proposed method, SAMWISE, achieves state-of-the-art across various benchmarks, by adding a negligible overhead of just 4.2 M parameters. The code is available at https://github.com/ClaudiaCuttano/SAMWISE

Recently, a class of mechanical lattices with reconfigurable, zero-stiffness structures has been proposed, called Totimorphic structures. In this work, we introduce a computational framework that allows continuous reprogramming of a Totimorphic lattice's effective properties, such as mechanical and optical properties, via continuous geometric changes alone. Our approach is differentiable and guarantees valid Totimorphic lattice configurations throughout the optimisation process, thus providing not only specific configurations with desired properties but also trajectories through configuration space connecting them. It enables re-programmable structures where actuators are controlled via automatic differentiation on an objective-dependent cost function, altering the lattice structure at all times to achieve a given objective - which is interchangeable to achieve different functionalities. Our main interest lies in deep space applications where harsh, extreme, and resource-constrained environments demand solutions that offer flexibility, resource efficiency, and autonomy. We illustrate our framework through two proofs of concept: a re-programmable metamaterial as well as a space telescope mirror with adjustable focal length, both made from Totimorphic structures. The introduced framework is easily adjustable to a variety of Totimorphic designs and objectives, providing a light-weight model for endowing physical prototypes of Totimorphic structures with autonomous self-configuration and self-repair capabilities.

Methyl cyanide is one of the simplest interstellar complex organic molecules, widely detected in young solar analogues, shocked regions, protoplanetary disks and comets. CH3CN can be considered a key species to explore the chemical connections between planet forming disks and comets. For such comparison to be meaningful kinetics data for the reactions leading to CH3CN formation and destruction must be updated. We focus on the destruction of methyl cyanide through collisions with He+. A combined experimental and theoretical methodology is employed to obtain cross sections (CSs) and branching ratios (BRs) as a function of collision energy, from which reaction rate coefficients $k(T)$ are calculated in the temperature range from 10 to 300 K. CSs and BRs are measured using a guided ion beam set-up. A theoretical treatment based on an analytical formulation of the potential energy surfaces (PESs) for the charge exchange process is developed. The method employs a Landau Zener model to obtain reaction probabilities at crossings between the entrance and exit PESs, and an adiabatic centrifugal sudden approximation to calculate CSs and k(T). Rates and BRs differ from those predicted from widely-used capture models. In particular, the rate coefficient at 10 K is estimated to be almost one order of magnitude smaller than what reported in the KIDA database. As for BRs, the charge exchange is completely dissociative and the most abundant fragments are HCCN+/CCNH+, HCNH+ and CH2+. Our results, combined with a revised chemical network for formation of CH3CN, support the hypothesis that methyl cyanide in protoplanetary disks could be mostly the product of gas-phase processes rather than grain chemistry, as currently proposed. These findings are expected to have implications in the comparison of the abundance ratios of N-bearing molecules observed in disks with cometary abundance ratios, Comment: 19 pages, 7 figures, full article

We present the largest optical photometry compilation of Gamma-Ray Bursts (GRBs) with redshifts ($z$). We include 64813 observations of 535 events (including upper limits) from 28 February 1997 up to 18 August 2023. We also present a user-friendly web tool \textit{grbLC} which allows users the visualization of photometry, coordinates, redshift, host galaxy extinction, and spectral indices for each event in our database. Furthermore, we have added a Gamma Ray Coordinate Network (GCN) scraper that can be used to collect data by gathering magnitudes from the GCNs. The web tool also includes a package for uniformly investigating colour evolution. We compute the optical spectral indices for 138 GRBs for which we have at least 4 filters at the same epoch in our sample and craft a procedure to distinguish between GRBs with and without colour evolution. By providing a uniform format and repository for the optical catalogue, this web-based archive is the first step towards unifying several community efforts to gather the photometric information for all GRBs with known redshifts. This catalogue will enable population studies by providing light curves (LCs) with better coverage since we have gathered data from different ground-based locations. Consequently, these LCs can be used to train future LC reconstructions for an extended inference of the redshift. The data gathering also allows us to fill some of the orbital gaps from Swift in crucial points of the LCs, e.g., at the end of the plateau emission or where a jet break is identified., Comment: 20 pages, 16 figures, 2 tables. Submitted to MNRAS, this version matches the third revision. The Online Materials and data will be available after the publication

Humans show an innate capability to identify tools to support specific actions. The association between objects parts and the actions they facilitate is usually named affordance. Being able to segment objects parts depending on the tasks they afford is crucial to enable intelligent robots to use objects of daily living. Traditional supervised learning methods for affordance segmentation require costly pixel-level annotations, while weakly supervised approaches, though less demanding, still rely on object-interaction examples and support a closed set of actions. These limitations hinder scalability, may introduce biases, and usually restrict models to a limited set of predefined actions. This paper proposes AffordanceCLIP, to overcome these limitations by leveraging the implicit affordance knowledge embedded within large pre-trained Vision-Language models like CLIP. We experimentally demonstrate that CLIP, although not explicitly trained for affordances detection, retains valuable information for the task. Our AffordanceCLIP achieves competitive zero-shot performance compared to methods with specialized training, while offering several advantages: i) it works with any action prompt, not just a predefined set; ii) it requires training only a small number of additional parameters compared to existing solutions and iii) eliminates the need for direct supervision on action-object pairs, opening new perspectives for functionality-based reasoning of models., Comment: Accepted to MAR Workshop at CVPR2024

Recent advancements in image segmentation have focused on enhancing the efficiency of the models to meet the demands of real-time applications, especially on edge devices. However, existing research has primarily concentrated on single-task settings, especially on semantic segmentation, leading to redundant efforts and specialized architectures for different tasks. To address this limitation, we propose a novel architecture for efficient multi-task image segmentation, capable of handling various segmentation tasks without sacrificing efficiency or accuracy. We introduce BiSeNetFormer, that leverages the efficiency of two-stream semantic segmentation architectures and it extends them into a mask classification framework. Our approach maintains the efficient spatial and context paths to capture detailed and semantic information, respectively, while leveraging an efficient transformed-based segmentation head that computes the binary masks and class probabilities. By seamlessly supporting multiple tasks, namely semantic and panoptic segmentation, BiSeNetFormer offers a versatile solution for multi-task segmentation. We evaluate our approach on popular datasets, Cityscapes and ADE20K, demonstrating impressive inference speeds while maintaining competitive accuracy compared to state-of-the-art architectures. Our results indicate that BiSeNetFormer represents a significant advancement towards fast, efficient, and multi-task segmentation networks, bridging the gap between model efficiency and task adaptability., Comment: Accepted to ECV workshop at CVPR2024

We present a comprehensive review on the state-of-the-art of the approximate analytic approaches describing the finite-temperature thermodynamic quantities of the Lieb-Liniger model of the one-dimensional (1D) Bose gas with contact repulsive interactions. This paradigmatic model of quantum many-body-theory plays an important role in many areas of physics -- thanks to its integrability and possible experimental realization using, e.g., ensembles of ultracold bosonic atoms confined to quasi-1D geometries. The thermodynamics of the uniform Lieb-Liniger gas can be obtained numerically using the exact thermal Bethe ansatz (TBA) method, first derived in 1969 by Yang and Yang. However, the TBA numerical calculations do not allow for the in-depth understanding of the underlying physical mechanisms that govern the thermodynamic behavior of the Lieb-Liniger gas at finite temperature. Our work is then motivated by the insights that emerge naturally from the transparency of closed-form analytic results, which are derived here in six different regimes of the gas and which exhibit an excellent agreement with the TBA numerics. Our findings can be further adopted for characterising the equilibrium properties of inhomogeneous (e.g., harmonically trapped) 1D Bose gases within the local density approximation and for the development of improved hydrodynamic theories, allowing for the calculation of breathing mode frequencies which depend on the underlying thermodynamic equation of state. Our analytic approaches can be applied to other systems including impurities in a quantum bath, liquid helium-4, and ultracold Bose gas mixtures., Comment: 55 pages, 9 figures; minor revisions in v2 in response to referee's comments

Recent transformer-based architectures have shown impressive results in the field of image segmentation. Thanks to their flexibility, they obtain outstanding performance in multiple segmentation tasks, such as semantic and panoptic, under a single unified framework. To achieve such impressive performance, these architectures employ intensive operations and require substantial computational resources, which are often not available, especially on edge devices. To fill this gap, we propose Prototype-based Efficient MaskFormer (PEM), an efficient transformer-based architecture that can operate in multiple segmentation tasks. PEM proposes a novel prototype-based cross-attention which leverages the redundancy of visual features to restrict the computation and improve the efficiency without harming the performance. In addition, PEM introduces an efficient multi-scale feature pyramid network, capable of extracting features that have high semantic content in an efficient way, thanks to the combination of deformable convolutions and context-based self-modulation. We benchmark the proposed PEM architecture on two tasks, semantic and panoptic segmentation, evaluated on two different datasets, Cityscapes and ADE20K. PEM demonstrates outstanding performance on every task and dataset, outperforming task-specific architectures while being comparable and even better than computationally-expensive baselines., Comment: CVPR 2024. Project page: https://niccolocavagnero.github.io/PEM

By means of numerical simulations, we demonstrate the innovative use of computational ghost imaging in transmission electron microscopy to retrieve images with a resolution that overcomes the limitations imposed by coherent aberrations. The method requires measuring the intensity on a single pixel detector with a series of structured illuminations. The success of the technique is improved if the probes are made to resemble the sample and the patterns cover the area of interest evenly. By using a simple 8 electrode device as a specific example, a 2-fold increase in resolution beyond the aberration limit is demonstrated to be possible under realistic experimental conditions., Comment: 8 pages, 3 figures

Methyl cyanide (CH$_3$CN) is one of the most abundant and widely spread interstellar complex organic molecules (iCOMs). Several studies found that, in hot corinos, methyl cyanide and methanol abundances are correlated suggesting a chemical link, often interpreted as a synthesis of them on the interstellar grain surfaces. In this article, we present a revised network of the reactions forming methyl cyanide in the gas-phase. We carried out an exhaustive review of the gas-phase CH$_3$CN formation routes, propose two new reactions and performed new quantum mechanics computations of several reactions. We found that 13 of the 15 reactions reported in the databases KIDA and UDfA have incorrect products and/or rate constants. The new corrected reaction network contains 10 reactions leading to methyl cyanide. We tested the relative importance of those reactions in forming CH$_3$CN using our astrochemical model. We confirm that the radiative association of CH${_3}{^+}$ and HCN, forming CH$_{3}$CNH$^{+}$, followed by the electron recombination of CH$_{3}$CNH$^{+}$, is the most important CH$_3$CN formation route in both cold and warm environments, notwithstanding that we significantly corrected the rate constants and products of both reactions. The two newly proposed reactions play an important role in warm environments. Finally, we found a very good agreement between the CH$_3$CN predicted abundances with those measured in cold ($\sim$10 K) and warm ($\sim$90 K) objects. Unexpectedly, we also found a chemical link between methanol and methyl cyanide via the CH$_{3}^{+}$ ion, which can explain the observed correlation between the CH$_3$OH and CH$_3$CN abundances measured in hot corinos., Comment: 24 pages, 19 figures, accepted in MNRAS

This paper introduces BWSNet, a model that can be trained from raw human judgements obtained through a Best-Worst scaling (BWS) experiment. It maps sound samples into an embedded space that represents the perception of a studied attribute. To this end, we propose a set of cost functions and constraints, interpreting trial-wise ordinal relations as distance comparisons in a metric learning task. We tested our proposal on data from two BWS studies investigating the perception of speech social attitudes and timbral qualities. For both datasets, our results show that the structure of the latent space is faithful to human judgements.

Positronium (Ps) has emerged as a promising test particle within the QUPLAS collaboration for investigating the gravitational effect. In this work, we present a novel approach to generate a monoenergetic and highly coherent Ps beam by creating a negative Ps ion (Ps$^-$, consisting of two electrons and one positron). The necessary positron beam is formed by using a high flux electron LINAC. Subsequently, we utilize a Fabry-Perot IR laser cavity operating at a wavelength of 1560 nm to selectively remove the extra electron. An alternative pulsed laser operating at a 3600 nm wavelength was studied to reduce broadening due to recoil and excitation. Here, we provide a Monte Carlo simulation to estimate the characteristics of the Ps beam, including its energy distribution and intensity profiles. The results obtained from this study will provide essential groundwork for future advancements in fundamental studies as Ps gravity measurements by using a Mach-Zehnder interferometer., Comment: 16 pages, 4 figures

This paper reports the study of a new interferometric configuration to measure the effect of gravity on positronium. A Mach-Zehnder matter-wave interferometer has been designed to operate with single-photon transitions and to transfer high momentum to a 200 eV positronium beam. The work shows the results and methods used to simulate the interferometer and estimate the operating parameters and the time needed to perform the experiment. It has been estimated that within less than one year, the acquisition time is sufficient to achieve a 10\% accuracy level in measuring positronium gravitational acceleration, even with a poorly collimated beam, which is significant for theoretical models describing matter-antimatter symmetry. These results pave the way for single photon transition large momentum transfer interferometry with fast atomic beams, which is particularly useful for studies with antimatter and unstable atoms., Comment: 24 pages, 11 figures

In triple-GEM detectors, the segmentation of GEM foils in electrically independent sectors allows reducing the probability of discharge damage to the detector and improving the detector rate capability; however, a segmented foil presents thin dead regions in the separation between two sectors and the segmentation pattern has to be manually aligned with the GEM hole pattern during the foil manufacturing, a procedure potentially sensitive to errors. We describe the production and characterization of triple-GEM detectors produced with an innovative GEM foil segmentation technique, the ``random hole segmentation'', that allows an easier manufacturing of segmented GEM foils. The electrical stability to high voltage and the gain uniformity of a random-hole segmented triple-GEM prototype are measured. The results of a test beam on a prototype assembled for the Phase-2 GEM upgrade of the CMS experiment are also presented; a high-statistics efficiency measurement shows that the random hole segmentation can limit the efficiency loss of the detector in the areas between two sectors, making it a viable alternative to blank segmentation for the GEM foil manufacturing of large-area detector systems.

GRB 221009A is the brightest Gamma-Ray Burst (GRB) detected in more than 50 years of study. In this paper, we present observations in the X-ray and optical domains after the GRB obtained by the GRANDMA Collaboration (which includes observations from more than 30 professional and amateur telescopes) and the Insight-HXMT Collaboration. We study the optical afterglow with empirical fitting from GRANDMA+HXMT data, augmented with data from the literature up to 60 days. We then model numerically, using a Bayesian approach, the GRANDMA and HXMT-LE afterglow observations, that we augment with Swift-XRT and additional optical/NIR observations reported in the literature. We find that the GRB afterglow, extinguished by a large dust column, is most likely behind a combination of a large Milky-Way dust column combined with moderate low-metallicity dust in the host galaxy. Using the GRANDMA+HXMT-LE+XRT dataset, we find that the simplest model, where the observed afterglow is produced by synchrotron radiation at the forward external shock during the deceleration of a top-hat relativistic jet by a uniform medium, fits the multi-wavelength observations only moderately well, with a tension between the observed temporal and spectral evolution. This tension is confirmed when using the extended dataset. We find that the consideration of a jet structure (Gaussian or power-law), the inclusion of synchrotron self-Compton emission, or the presence of an underlying supernova do not improve the predictions, showing that the modelling of GRB22109A will require going beyond the most standard GRB afterglow model. Placed in the global context of GRB optical afterglows, we find the afterglow of GRB 221009A is luminous but not extraordinarily so, highlighting that some aspects of this GRB do not deviate from the global known sample despite its extreme energetics and the peculiar afterglow evolution., Comment: Accepted to ApJL for the special issue, 37 pages, 23 pages main text, 6 tables, 13 figures

We study the thermal behavior of correlations in a one-dimensional Bose gas with tunable interaction strength, crossing from weakly-repulsive to Tonks-Girardeau regime. A reference temperature in this system is that of the hole anomaly, observed as a peak in the specific heat and a maximum in the chemical potential. We find that at large momenta $k$ and temperature above the anomaly threshold, the tail $\mathcal{C}/k^4$ of the momentum distribution (proportional to the Tan contact $\mathcal{C}$) is screened by the $1/|k|^3$-term due to a dramatic thermal increase of the internal energy emerging from the thermal occupation of spectral excitation states. The same fading is consistently revealed in the behavior at short distances $x$ of the one-body density matrix (OBDM) where the $|x|^3$-dependence disappears for temperatures above the anomaly. We obtain a new general analytic tail for the momentum distribution and a minimum $k$ fixing its validity range, both calculated with exact Bethe-Ansatz method and valid in all interaction and thermal regimes, crossing from the quantum to the classical gas limit. Our predictions are confirmed by comparison with ab-initio Path Integral Monte Carlo calculations for the momentum distribution and the OBDM exploring a wide range of interaction strength and temperature. Our results unveil a novel connection between excitations and correlations. We expect them to be of interest to any cold atomic, nuclear, solid-state, electronic and spin system exhibiting an anomaly or a thermal second-order phase transition., Comment: Main Text: 7 pages, 4 figures; Supplemental Material: 10 pages, 8 figures

We present a comprehensive study shedding light on how thermal fluctuations affect correlations in a Bose gas with contact repulsive interactions in one spatial dimension. The pair correlation function, the static structure factor, and the one-body density matrix are calculated as a function of the interaction strength and temperature with the exact ab-initio Path Integral Monte Carlo method. We explore all possible gas regimes from weak to strong interactions and from low to high temperatures. We provide a detailed comparison with a number of theories, such as perturbative (Bogoliubov and decoherent classical), effective (Luttinger liquid) and exact (ground-state and thermal Bethe Ansatz) ones. Our Monte Carlo results exhibit an excellent agreement with the tractable limits and provide a fundamental benchmark for future observations which can be achieved in atomic gases, cavity quantum-electrodynamic and superconducting-circuit platforms., Comment: Main Text: 13 pages, 9 figures. Appendix: 1 page