Search

Convolutional neural networks (CNNs) have been pivotal in various 2D image analysis tasks, including computer vision, image indexing and retrieval or semantic classification. Extending CNNs to 3D data such as point clouds and 3D meshes raises significant challenges since the very basic convolution and pooling operators need to be completely re-visited and re-defined in an appropriate manner to tackle irregular connectivity issues. In this paper, we introduce MeshConv3D, a 3D mesh-dedicated methodology integrating specialized convolution and face collapse-based pooling operators. MeshConv3D operates directly on meshes of arbitrary topology, without any need of prior re-meshing/conversion techniques. In order to validate our approach, we have considered a semantic classification task. The experimental results obtained on three distinct benchmark datasets show that the proposed approach makes it possible to achieve equivalent or superior classification results, while minimizing the related memory footprint and computational load.

We consider the kinetic wave equation, or phonon Boltzmann equation, set on the torus (physical system set on the lattice). We describe entropy maximizers for fixed mass and energy; our framework is very general, being valid in any dimension, for any dispersion relation, and even including the quantum kinetic wave equation. Of particular interest is the presence of condensation in certain regimes which we characterize., Comment: 10 pages, 1 figure

The iterative bleaching extends multiplexity (IBEX) Knowledge-Base is a central portal for researchers adopting IBEX and related 2D and 3D immunofluorescence imaging methods. The design of the Knowledge-Base is modeled after efforts in the open-source software community and includes three facets: a development platform (GitHub), static website, and service for data archiving. The Knowledge-Base facilitates the practice of open science throughout the research life cycle by providing validation data for recommended and non-recommended reagents, e.g., primary and secondary antibodies. In addition to reporting negative data, the Knowledge-Base empowers method adoption and evolution by providing a venue for sharing protocols, videos, datasets, software, and publications. A dedicated discussion forum fosters a sense of community among researchers while addressing questions not covered in published manuscripts. Together, scientists from around the world are advancing scientific discovery at a faster pace, reducing wasted time and effort, and instilling greater confidence in the resulting data.

The photo-electric effect was a historic milestone in the development of quantum theory, revealing the first evidence of discrete energy of the electromagnetic field through hallmark signatures such as the threshold frequency, intensity-independent energy transfer, and the near instantaneous ejection of photo-electrons. In this work, we revisit the photo-electric effect through the lens of semi-classical, quantum, and neo-classical models to explore their implications on energy conservation and discrete photon exchange. We extend this analysis to include the gravito-phononic effect through modelling energy transfer between coupled harmonic oscillators, rather than traditional atom-field interactions. Our approach demonstrates that two coupled harmonic oscillators, under a beam-splitter interaction, can exhibit hallmark signatures analogous to the photo-electric effect, including resonance conditions and quantised energy absorption. We discuss the implications of this model for the recently proposed gravito-phononic detections of single gravitons. This further clarifies that a gravitational version of the photo-electric effect, modelled as discrete energy transfer between harmonic oscillators can provide the first evidence of the graviton., Comment: 5 + 1 pages, 1 figure. A short note answering common questions

The monotone path polytope of a polytope $P$ encapsulates the combinatorial behavior of the shadow vertex rule (a pivot rule used in linear programming) on $P$. On the other hand, computing monotone path polytopes is the entry door to the larger subject of fiber polytopes, for which explicitly computing examples remains a challenge. Monotone path polytopes of cubes and simplices have been known since the seminal article of Billera and Strumfels 1992. We (partially) extend these results to hypersimplices by linking this problem to the combinatorics of lattice paths. Indeed, we give a combinatorial model to describe the vertices of the monotone path polytope of the hypersimplex $\Delta(n, 2)$ (for any generic direction). With this model, we give a precise count of these vertices, and furthermore count the number of coherent monotone paths on $\Delta(n, 2)$ according to their lengths. We prove that some of the results obtained also hold for hypersimplices $\Delta(n, k)$ for $k\geq 2$., Comment: 25 pages, 14 figures

This document summarizes the prospective physics plans of the French QCD and Heavy-Ion community, including the experimental programs at the LHC Run 5 and beyond and future colliders at CERN, within the context of the French contribution to the update of the European Strategy in Particle Physics (ESPPU 2025), as discussed in the workshop on European Strategy for Particle Physics Update 2025 organised by the QCD GdR in Oct. 2024., Comment: 6.5 pages without title page and without references, no figures. Keywords: QCD, Heavy Ions, Quark Gluon Plasma, LHC, FCC

Several attempts have been made to handle multiple source separation tasks such as speech enhancement, speech separation, sound event separation, music source separation (MSS), or cinematic audio source separation (CASS) with a single model. These models are trained on large-scale data including speech, instruments, or sound events and can often successfully separate a wide range of sources. However, it is still challenging for such models to cover all separation tasks because some of them are contradictory (e.g., musical instruments are separated in MSS while they have to be grouped in CASS). To overcome this issue and support all the major separation tasks, we propose a task-aware unified source separation (TUSS) model. The model uses a variable number of learnable prompts to specify which source to separate, and changes its behavior depending on the given prompts, enabling it to handle all the major separation tasks including contradictory ones. Experimental results demonstrate that the proposed TUSS model successfully handles the five major separation tasks mentioned earlier. We also provide some audio examples, including both synthetic mixtures and real recordings, to demonstrate how flexibly the TUSS model changes its behavior at inference depending on the prompts., Comment: Submitted to ICASSP 2025

Reconstruction functions are pivotal in sample compression theory, a framework for deriving tight generalization bounds. From a small sample of the training set (the compression set) and an optional stream of information (the message), they recover a predictor previously learned from the whole training set. While usually fixed, we propose to learn reconstruction functions. To facilitate the optimization and increase the expressiveness of the message, we derive a new sample compression generalization bound for real-valued messages. From this theoretical analysis, we then present a new hypernetwork architecture that outputs predictors with tight generalization guarantees when trained using an original meta-learning framework. The results of promising preliminary experiments are then reported., Comment: Accepted at the NeurIPS 2024 workshop on Compression in Machine Learning

We provide a piecewise linear isomorphism from the normal fan of the pivot polytope of a product of simplices to the normal fan of a shuffle of associahedra., Comment: 7 pages, 1 figure

Two-dimensional Euler flows, in the plane or on simple surfaces, possess a material invariant, namely the scalar vorticity normal to the surface. Consequently, flows with piecewise-uniform vorticity remain that way, and moreover evolve in a way which is entirely determined by the instantaneous shapes of the contours (interfaces) separating different regions of vorticity -- this is known as `Contour Dynamics'. Unsteady vorticity contours or interfaces often grow in complexity (lengthen and fold), either as a result of vortex interactions (like merger) or `filamentation'. In the latter, wave disturbances riding on a background, equilibrium contour shape appear to inevitably steepen and break, forming filaments, repeatedly -- and perhaps endlessly. Here, we revisit the onset of filamentation. Building upon previous work and using a weakly-nonlinear expansion to third order in wave amplitude, we derive a universal, parameter-free amplitude equation which applies (with a minor change) both to a straight interface and a circular patch in the plane, as well as circular vortex patches on the surface of a sphere. We show that this equation possesses a local, self-similar form describing the finite-time blow up of the wave slope (in a rescaled long time proportional to the inverse square of the initial wave amplitude). We present numerical evidence for this self-similar blow-up solution, and for the conjecture that almost all initial conditions lead to finite-time blow up. In the full Contour Dynamics equations, this corresponds to the onset of filamentation., Comment: 19 pages, 15 figures

We study an asymptotic nonlinear model for filamention on two-dimensional vorticity interfaces. Different re-formulations of the model equation reveal its underlying structural properties. They enable us to construct global weak solutions and to prove the existence of traveling waves., Comment: 12 pages

We review asymptotic stability of solitary waves for nonlinear dispersive equations set on the line. Our focus is threefold: first, the nonlinear Schrodinger equation; second, the notion of full asymptotic stability (which states that perturbations of a solitary wave decompose globally into a solitary wave and a decaying solution); and third, spectral methods. Besides this focus, we summarize the state of the art in a broader context, including nonlinear Klein-Gordon equations, the notion of local asymptotic stability, and virial methods., Comment: 19 pages, 5 figures

We present a foundation model for zero-shot metric monocular depth estimation. Our model, Depth Pro, synthesizes high-resolution depth maps with unparalleled sharpness and high-frequency details. The predictions are metric, with absolute scale, without relying on the availability of metadata such as camera intrinsics. And the model is fast, producing a 2.25-megapixel depth map in 0.3 seconds on a standard GPU. These characteristics are enabled by a number of technical contributions, including an efficient multi-scale vision transformer for dense prediction, a training protocol that combines real and synthetic datasets to achieve high metric accuracy alongside fine boundary tracing, dedicated evaluation metrics for boundary accuracy in estimated depth maps, and state-of-the-art focal length estimation from a single image. Extensive experiments analyze specific design choices and demonstrate that Depth Pro outperforms prior work along multiple dimensions. We release code and weights at https://github.com/apple/ml-depth-pro, Comment: Code and weights available at https://github.com/apple/ml-depth-pro

Despite the prevalence and historical importance of maps in digital collections, current methods of navigating and exploring map collections are largely restricted to catalog records and structured metadata. In this paper, we explore the potential for interactively searching large-scale map collections using natural language inputs ("maps with sea monsters"), visual inputs (i.e., reverse image search), and multimodal inputs (an example map + "more grayscale"). As a case study, we adopt 562,842 images of maps publicly accessible via the Library of Congress's API. To accomplish this, we use the mulitmodal Contrastive Language-Image Pre-training (CLIP) machine learning model to generate embeddings for these maps, and we develop code to implement exploratory search capabilities with these input strategies. We present results for example searches created in consultation with staff in the Library of Congress's Geography and Map Division and describe the strengths, weaknesses, and possibilities for these search queries. Moreover, we introduce a fine-tuning dataset of 10,504 map-caption pairs, along with an architecture for fine-tuning a CLIP model on this dataset. To facilitate re-use, we provide all of our code in documented, interactive Jupyter notebooks and place all code into the public domain. Lastly, we discuss the opportunities and challenges for applying these approaches across both digitized and born-digital collections held by galleries, libraries, archives, and museums., Comment: 18 pages, 7 figures, accepted at the Computational Humanities Research Conference (CHR 2024)

The sample compression theory provides generalization guarantees for predictors that can be fully defined using a subset of the training dataset and a (short) message string, generally defined as a binary sequence. Previous works provided generalization bounds for the zero-one loss, which is restrictive notably when applied to deep learning approaches. In this paper, we present a general framework for deriving new sample compression bounds that hold for real-valued unbounded losses. Using the Pick-To-Learn (P2L) meta-algorithm, which transforms the training method of any machine-learning predictor to yield sample-compressed predictors, we empirically demonstrate the tightness of the bounds and their versatility by evaluating them on random forests and multiple types of neural networks.

Effective field theory descriptions of surface waves on flowing fluids have tended to assume that the flow is irrotational, but this assumption is often impractical due to boundary layer friction and flow recirculation. Here we develop an effective field theory of surface waves in an incompressible, inviscid flow that includes vorticity due to shear. Our model consists of a two-layer flow: an upper layer with no vorticity and a lower layer with constant vorticity. We consider linear, long-wavelength perturbations on top of such a flow, and find that these can be described by two coupled scalar fields admitting three elementary excitations, one more than the usual two found in irrotational flows. We compute the scattering coefficients pertaining to modes falling into an analogue black hole. Our approach provides a more realistic framework for simulating gravitational wave phenomena possibly with an internal structure mimicking quantum gravity effects in laboratory settings.

The goal of text-queried target sound extraction (TSE) is to extract from a mixture a sound source specified with a natural-language caption. While it is preferable to have access to large-scale text-audio pairs to address a variety of text prompts, the limited number of available high-quality text-audio pairs hinders the data scaling. To this end, this work explores how to leverage audio-only data without any captions for the text-queried TSE task to potentially scale up the data amount. A straightforward way to do so is to use a joint audio-text embedding model, such as the contrastive language-audio pre-training (CLAP) model, as a query encoder and train a TSE model using audio embeddings obtained from the ground-truth audio. The TSE model can then accept text queries at inference time by switching to the text encoder. While this approach should work if the audio and text embedding spaces in CLAP were well aligned, in practice, the embeddings have domain-specific information that causes the TSE model to overfit to audio queries. We investigate several methods to avoid overfitting and show that simple embedding-manipulation methods such as dropout can effectively alleviate this issue. Extensive experiments demonstrate that using audio-only data with embedding dropout is as effective as using text captions during training, and audio-only data can be effectively leveraged to improve text-queried TSE models., Comment: Submitted to ICASSP 2025

Deep learning models have been shown to be a powerful solution for Time Series Classification (TSC). State-of-the-art architectures, while producing promising results on the UCR and the UEA archives , present a high number of trainable parameters. This can lead to long training with high CO2 emission, power consumption and possible increase in the number of FLoating-point Operation Per Second (FLOPS). In this paper, we present a new architecture for TSC, the Light Inception with boosTing tEchnique (LITE) with only 2.34% of the number of parameters of the state-of-the-art InceptionTime model, while preserving performance. This architecture, with only 9, 814 trainable parameters due to the usage of DepthWise Separable Convolutions (DWSC), is boosted by three techniques: multiplexing, custom filters, and dilated convolution. The LITE architecture, trained on the UCR, is 2.78 times faster than InceptionTime and consumes 2.79 times less CO2 and power. To evaluate the performance of the proposed architecture on multivariate time series data, we adapt LITE to handle multivariate time series, we call this version LITEMV. To bring theory into application, we also conducted experiments using LITEMV on multivariate time series representing human rehabilitation movements, showing that LITEMV not only is the most efficient model but also the best performing for this application on the Kimore dataset, a skeleton based human rehabilitation exercises dataset. Moreover, to address the interpretability of LITEMV, we present a study using Class Activation Maps to understand the classification decision taken by the model during evaluation.

We study the nonlinear dynamics of the kinetic wave equation associated to the FPU problem and prove stability of the non-singular Rayleigh-Jeans equilibria. The lack of a spectral gap for the linearized problem leads to polynomial decay, which we are able to leverage to obtain nonlinear stability., Comment: 34 pages

Accelerating/decelerating trans-critical flows (waterfalls/cataracts) are analogous to space-times of black holes/white fountains since the pioneering work of Sch{\"u}tzhold \& Unruh in 2002. A single number is usually employed to classify trans-criticality namely the local depth Froude number which is the ratio between the local current speed and the local celerity of long gravity waves analogous to the light celerity. When the former reaches one, water waves are no more able to propagate upstream: the hydraulic black hole is a river of no return for them. At a higher level of understanding, two global dimensionless numbers, the upstream Froude number F r up and the obstruction ratio r up (the height of a bottom obstacle, the underlying geometry inducing the effective space-time, divided by the upstream water depth) are essential to distinguish subcritical, trans-critical and supercritical zones in the -F r up versus r up -hydraulic and nondispersive diagram. The relationship between both global parameters for transcritical flows turns out to be a peculiar limit of the behaviour of boats navigating in confined media like canals or locks with a generalized obstruction factor based on the ratio between the boat section and the canal section. Here, we revisit the classification of flows over obstacles in open water channel taking into account both effects of dispersion and scale, two neglected topics so far. For the first time, we give a complete classification of flows in an open water channel based on sub-pixel detection method measurements of the free surface supported by numerical simulations. We generalized the obstruction factor by a filling factor taking into account the maximum height of the water channel, a crucial parameter that was overlooked so far. Our ultimate purpose is to understand how to reproduce in the laboratory analogues of curved space-times from the dynamical point of view.

Second-order Raman scattering has been extensively studied in carbon-based nanomaterials, \emph{e.g.} nanotube and graphene, because it activates normally forbidden Raman modes that are sensitive to crystal disorder, such as defects, dopants, strain, etc. The sp$^2$-hybridized carbon systems are, however, the exception among most nanomaterials, where first-order Raman processes usually dominate. Here we report the identification of four second-order Raman modes, named $D_1$, $D_1'$, $D_2$ and $D_2'$, in exfoliated black phosphorus (P(black)), an elemental direct-gap semiconductor exhibiting strong mechanical and electronic anisotropies. Located in close proximity to the $A^1_g$ and $A^2_g$ modes, these new modes dominate at an excitation wavelength of 633 nm. Their evolutions as a function of sample thickness, excitation wavelength, and defect density indicate that they are defect-activated and involve high-momentum phonons in a doubly-resonant Raman process. \emph{Ab initio} simulations of a monolayer reveal that the $D'$ and $D$ modes occur through intravalley scatterings with split contributions in the armchair and zigzag directions, respectively. The high sensitivity of these $D$ modes to disorder helps explaining several discrepancies found in the literature., Comment: 18 pages, 5 figures, with supplementary file

Reverberation as supervision (RAS) is a framework that allows for training monaural speech separation models from multi-channel mixtures in an unsupervised manner. In RAS, models are trained so that sources predicted from a mixture at an input channel can be mapped to reconstruct a mixture at a target channel. However, stable unsupervised training has so far only been achieved in over-determined source-channel conditions, leaving the key determined case unsolved. This work proposes enhanced RAS (ERAS) for solving this problem. Through qualitative analysis, we found that stable training can be achieved by leveraging the loss term to alleviate the frequency-permutation problem. Separation performance is also boosted by adding a novel loss term where separated signals mapped back to their own input mixture are used as pseudo-targets for the signals separated from other channels and mapped to the same channel. Experimental results demonstrate high stability and performance of ERAS., Comment: Accepted to Interspeech 2024

Time-frequency (TF) domain dual-path models achieve high-fidelity speech separation. While some previous state-of-the-art (SoTA) models rely on RNNs, this reliance means they lack the parallelizability, scalability, and versatility of Transformer blocks. Given the wide-ranging success of pure Transformer-based architectures in other fields, in this work we focus on removing the RNN from TF-domain dual-path models, while maintaining SoTA performance. This work presents TF-Locoformer, a Transformer-based model with LOcal-modeling by COnvolution. The model uses feed-forward networks (FFNs) with convolution layers, instead of linear layers, to capture local information, letting the self-attention focus on capturing global patterns. We place two such FFNs before and after self-attention to enhance the local-modeling capability. We also introduce a novel normalization for TF-domain dual-path models. Experiments on separation and enhancement datasets show that the proposed model meets or exceeds SoTA in multiple benchmarks with an RNN-free architecture., Comment: Accepted to IWAENC 2024

Iterative Bleaching Extends multipleXity (IBEX) is a versatile method for highly multiplexed imaging of diverse tissues. Based on open science principles, we created the IBEX Knowledge-Base, a resource for reagents, protocols and more, to empower innovation., Comment: 8 pages, 1 figure, 9 references

To leverage the angular resolution of interferometry at high contrast, one must employ specialized beam-combiners called interferometric nullers. Nullers discard part of the astrophysical information to optimize the recording of light present in the dark fringe of the central source. Asgard/NOTT will deploy a beam-combination scheme offering good instrumental noise rejection when phased appropriately, but for which information is degenerate on the outputs, prompting a dedicated tuning strategy using the science detector. The dispersive effect of water vapor can be corrected with prisms forming a variable thickness of glass. But observations in the L band suffer from an additional and important chromatic effect due to longitudinal atmospheric dispersion coming from a resonance of CO2 at 4.3 micron. To compensate for this effect efficiently, a novel type of compensation device will be deployed leveraging a gas cell of variable length at ambient pressure. After reviewing the impact of water vapor and CO2, we present the design of this atmospheric dispersion compensation device and describe a strategy to maintain this tuning on-sky., Comment: 7 pages, 5 figures, Proc of SPIE Astronomical Telescopes + Instrumentation 2024 Yokohama, Japan, Optical and Infrared Interferometry and Imaging VI

In this note, we show that the sequence of maximum values in frieze patterns of type $A_n$ is the sequence of Fibonacci numbers, and that of frieze patterns of type $C_n$ is the sequence of odd Fibonacci numbers., Comment: Latest version with upper bounds in types An and Cn. Note associated to a Lean4 formalisation project available on GitHub

In a recent work we showed that the detection of the exchange of a single graviton between a massive quantum resonator and a gravitational wave can be achieved. Key to this ability are the experimental progress in preparing and measuring massive resonators in the quantum regime, and the correlation with independent LIGO detections of gravitational waves that induce stimulated absorption. But do stimulated single-graviton processes imply the quantization of gravity? Here we analyze this question and make a historic analogy to the early days of quantum theory. We discuss in what ways such experiments can indeed probe key features of the quantized interaction between gravity and matter, and outline five experimental tests. This capability would open the first window into experimental exploration of quantum gravity., Comment: 10 pages + references, Based on essay selected for Honorable Mention by the Gravity Research Foundation 2024 Awards for Essays on Gravitation

Nulling interferometry is a powerful observing technique to reach exoplanets and circumstellar dust at separations too small for direct imaging with single-dish telescopes and too large for indirect methods. With near-future instrumentation, it bears the potential to detect young, hot planets near the snow lines of their host stars. A future space mission could detect and characterize a large number of rocky, habitable-zone planets around nearby stars at thermal-infrared wavelengths. The null self-calibration is a method aiming at modelling the statistical distribution of the nulled signal. It has proven to be more sensitive and accurate than average-based data reduction methods in nulling interferometry. This statistical approach opens the possibility of designing a GPU-based Python package to reduce the data from any of these instruments, by simply providing the data and a simulator of the instrument. GRIP is a toolbox to reduce nulling and interferometric data based on the statistical self-calibration method. In this article, we present the main features of GRIP as well as applications on real data.

ESO's Very Large Telescope Interferometer has a history of record-breaking discoveries in astrophysics and significant advances in instrumentation. The next leap forward is its new visitor instrument, called Asgard. It comprises four natively collaborating instruments: HEIMDALLR, an instrument performing both fringe tracking and stellar interferometry simultaneously with the same optics, operating in the K band; Baldr, a Strehl optimizer in the H band; BIFROST, a spectroscopic combiner to study the formation processes and properties of stellar and planetary systems in the Y-J-H bands; and NOTT, a nulling interferometer dedicated to imaging nearby young planetary systems in the L band. The suite is in its integration phase in Europe and should be shipped to Paranal in 2025. In this article, we present details of the alignment and calibration unit, the observing modes, the integration plan, the software architecture, and the roadmap to completion of the project.

NOTT (formerly Hi-5) is the L'-band (3.5-4.0~microns) nulling interferometer of Asgard, an instrument suite in preparation for the VLTI visitor focus. The primary scientific objectives of NOTT include characterizing (i) young planetary systems near the snow line, a critical region for giant planet formation, and (ii) nearby main-sequence stars close to the habitable zone, with a focus on detecting exozodiacal dust that could obscure Earth-like planets. In 2023-2024, the final warm optics have been procured and assembled in a new laboratory at KU Leuven. First fringes and null measurements were obtained using a Gallium Lanthanum Sulfide (GLS) photonic chip that was also tested at cryogenic temperatures. In this paper, we present an overall update of the NOTT project with a particular focus on the cold mechanical design, the first results in the laboratory with the final NOTT warm optics, and the ongoing Asgard integration activities. We also report on other ongoing activities such as the characterization of the photonic chip (GLS, LiNbO3, SiO), the development of the exoplanet science case, the design of the dispersion control module, and the progress with the self-calibration data reduction software., Comment: 11 pages (incl. 5 figures); Proc. SPIE Astronomical Telescopes + Instrumentation 2024 (Yokohama; Japan), Optical and Infrared Interferometry and Imaging VI

Consider an eigenfunction of the Laplacian on a torus. How small can its $L^2$-norm be on small balls? We provide partial answers to this question by exploiting the distribution of integer points on spheres, basic properties of polynomials, and Nazarov--Tur\'an type estimates for exponential polynomials. Applications to quantum limits and control theory are given., Comment: 18 pages

In this paper, we propose Phoneme Discretized Saliency Maps (PDSM), a discretization algorithm for saliency maps that takes advantage of phoneme boundaries for explainable detection of AI-generated voice. We experimentally show with two different Text-to-Speech systems (i.e., Tacotron2 and Fastspeech2) that the proposed algorithm produces saliency maps that result in more faithful explanations compared to standard posthoc explanation methods. Moreover, by associating the saliency maps to the phoneme representations, this methodology generates explanations that tend to be more understandable than standard saliency maps on magnitude spectrograms., Comment: Proc. Interspeech 2024, 3295-3299, doi: 10.21437/Interspeech.2024-632

Sound event detection is the task of recognizing sounds and determining their extent (onset/offset times) within an audio clip. Existing systems commonly predict sound presence confidence in short time frames. Then, thresholding produces binary frame-level presence decisions, with the extent of individual events determined by merging consecutive positive frames. In this paper, we show that frame-level thresholding degrades the prediction of the event extent by coupling it with the system's sound presence confidence. We propose to decouple the prediction of event extent and confidence by introducing SEBBs, which format each sound event prediction as a tuple of a class type, extent, and overall confidence. We also propose a change-detection-based algorithm to convert legacy frame-level outputs into SEBBs. We find the algorithm significantly improves the performance of DCASE 2023 Challenge systems, boosting the state of the art from .644 to .686 PSDS1., Comment: Accepted for publication at Interspeech 2024

We report a theoretical and experimental investigation of fiber Fabry-Perot cavities aimed at enhancing Kerr frequency comb generation. The modulation instability (MI) power threshold is derived from the linear stability analysis of a generalized Lugiato-Lefever equation. By combining this analysis with the concepts of power enhancement factor (PEF) and optimal coupling, we predict the ideal manufacturing parameters of fiber Fabry-Perot (FFP) cavities for the MI Kerr frequency comb generation. Our findings reveal a distinction between the optimal coupling for modulation instability and that of the cold cavity. Consequently, mirror reflectivity must be adjusted to suit the specific application. We verified the predictions of our theory by measuring the MI power threshold as a function of detuning for three different cavities.

We propose a multi-mode bar consisting of mass elements of decreasing size for the implementation of a gravitational version of the photo-electric effect through the stimulated absorption of up to kHz gravitons from a binary neutron star merger and post-merger. We find that the multi-mode detector has normal modes that retain the coupling strength to the gravitational wave of the largest mass-element, while only having an effective mass comparable to the mass of the smallest element. This allows the normal modes to have graviton absorption rates due to the tonne-scale largest mass, while the single graviton absorption process in the normal mode could be resolved through energy measurements of a mass-element in-principle smaller than pico-gram scale. We argue the feasibility of directly counting gravito-phonons in the bar through energy measurements of the end mass. This improves the transduction of the single-graviton signal, enhancing the feasibility of detecting single gravitons., Comment: 8 + 3 pages, 2 Figures

Optimization layers in deep neural networks have enjoyed a growing popularity in structured learning, improving the state of the art on a variety of applications. Yet, these pipelines lack interpretability since they are made of two opaque layers: a highly non-linear prediction model, such as a deep neural network, and an optimization layer, which is typically a complex black-box solver. Our goal is to improve the transparency of such methods by providing counterfactual explanations. We build upon variational autoencoders a principled way of obtaining counterfactuals: working in the latent space leads to a natural notion of plausibility of explanations. We finally introduce a variant of the classic loss for VAE training that improves their performance in our specific structured context. These provide the foundations of CF-OPT, a first-order optimization algorithm that can find counterfactual explanations for a broad class of structured learning architectures. Our numerical results show that both close and plausible explanations can be obtained for problems from the recent literature.

The development of generative artificial intelligence for human motion generation has expanded rapidly, necessitating a unified evaluation framework. This paper presents a detailed review of eight evaluation metrics for human motion generation, highlighting their unique features and shortcomings. We propose standardized practices through a unified evaluation setup to facilitate consistent model comparisons. Additionally, we introduce a novel metric that assesses diversity in temporal distortion by analyzing warping diversity, thereby enhancing the evaluation of temporal data. We also conduct experimental analyses of three generative models using a publicly available dataset, offering insights into the interpretation of each metric in specific case scenarios. Our goal is to offer a clear, user-friendly evaluation framework for newcomers, complemented by publicly accessible code.

Quantum computing has the potential for disruptive change in many sectors of industry, especially in materials science and optimization. In this paper, we describe how the Turbine Balancing Problem can be solved with quantum computing, which is the NP-hard optimization problem of analytically balancing rotor blades in a single plane as found in turbine assembly. Small yet relevant instances occur in industry, which makes the problem interesting for early quantum computing benchmarks. We model it as a Quadratic Unconstrained Binary Optimization problem and compare the performance of a classical rule-based heuristic and D-Wave Systems' Quantum Annealer Advantage_system4.1. In this case study, we use real-world as well as synthetic datasets and observe that the quantum hardware significantly improves an actively used heuristic's solution for small-scale problem instances with bare disk imbalance in terms of solution quality. Motivated by this performance gain, we subsequently design a quantum-inspired classical heuristic based on simulated annealing that achieves extremely good results on all given problem instances, essentially solving the optimization problem sufficiently well for all considered datasets, according to industrial requirements., Comment: 6 pages, 3 figure

We study tropical friezes and cluster-additive functions associated to symmetrizable generalized Cartan matrices in the framework of Fock-Goncharov duality in cluster algebras. In particular, we generalize and prove a conjecture of C. M. Ringel on cluster-additive functions associated to arbitrary Cartan matrices of finite type. For a Fock-Goncharov dual pair of positive spaces of finite type, we use tropical friezes and cluster-additive functions to explicitly express the Fock-Goncharov pairing between their tropical points and the bijections between global monomials on one and tropical points of the other., Comment: intended as a replacement of arXiv:2311.17712v1

We give an elementary account of the notion of Y-frieze patterns, explain some of their properties, and reveal their connection with Coxeter's frieze patterns.