Search

This paper aims to develop the study of historical printed ornaments with modern unsupervised computer vision. We highlight three complex tasks that are of critical interest to book historians: clustering, element discovery, and unsupervised change localization. For each of these tasks, we introduce an evaluation benchmark, and we adapt and evaluate state-of-the-art models. Our Rey's Ornaments dataset is designed to be a representative example of a set of ornaments historians would be interested in. It focuses on an XVIIIth century bookseller, Marc-Michel Rey, providing a consistent set of ornaments with a wide diversity and representative challenges. Our results highlight the limitations of state-of-the-art models when faced with real data and show simple baselines such as k-means or congealing can outperform more sophisticated approaches on such data. Our dataset and code can be found at https://printed-ornaments.github.io/.

Representing and quantifying Minimal Residual Disease (MRD) in Acute Myeloid Leukemia (AML), a type of cancer that affects the blood and bone marrow, is essential in the prognosis and follow-up of AML patients. As traditional cytological analysis cannot detect leukemia cells below 5\%, the analysis of flow cytometry dataset is expected to provide more reliable results. In this paper, we explore statistical learning methods based on optimal transport (OT) to achieve a relevant low-dimensional representation of multi-patient flow cytometry measurements (FCM) datasets considered as high-dimensional probability distributions. Using the framework of OT, we justify the use of the K-means algorithm for dimensionality reduction of multiple large-scale point clouds through mean measure quantization by merging all the data into a single point cloud. After this quantization step, the visualization of the intra and inter-patients FCM variability is carried out by embedding low-dimensional quantized probability measures into a linear space using either Wasserstein Principal Component Analysis (PCA) through linearized OT or log-ratio PCA of compositional data. Using a publicly available FCM dataset and a FCM dataset from Bordeaux University Hospital, we demonstrate the benefits of our approach over the popular kernel mean embedding technique for statistical learning from multiple high-dimensional probability distributions. We also highlight the usefulness of our methodology for low-dimensional projection and clustering patient measurements according to their level of MRD in AML from FCM. In particular, our OT-based approach allows a relevant and informative two-dimensional representation of the results of the FlowSom algorithm, a state-of-the-art method for the detection of MRD in AML using multi-patient FCM.

Elastic wave manipulation using large arrays of resonators is driving the need for advanced simulation and optimization methods. To address this we introduce and explore a robust framework for wave control: Quasi-normal modes (QNMs). Specifically we consider the problem for thin elastic plates, where the Green's function formalism is well known and readily exploited to solve multiple scattering problems. By studying the associated nonlinear eigenvalue problem we derive a dispersive QNM expansion, providing a reduced-order model for efficient forced response computations which reveals physical insight into the resonant mode excitation. Furthermore, we derive eigenvalue sensitivities with respect to resonator parameters and apply a gradient-based optimization to design quasi-bound states in the continuum and position eigenfrequencies precisely in the complex plane. Scattering simulations validate our approach in structures such as graded line arrays and quasi-crystals. Drawing on QNM concepts from electromagnetism we demonstrate significant advances in elastic metamaterials, highlighting their potential for tailored wave manipulation.

This survey reviews the AIS 2024 Event-Based Eye Tracking (EET) Challenge. The task of the challenge focuses on processing eye movement recorded with event cameras and predicting the pupil center of the eye. The challenge emphasizes efficient eye tracking with event cameras to achieve good task accuracy and efficiency trade-off. During the challenge period, 38 participants registered for the Kaggle competition, and 8 teams submitted a challenge factsheet. The novel and diverse methods from the submitted factsheets are reviewed and analyzed in this survey to advance future event-based eye tracking research., Comment: Qinyu Chen is the corresponding author

The Chow rings of hyper-K\"ahler varieties are conjectured to have a particularly rich structure. In this paper, we formulate a conjecture that combines the Beauville-Voisin conjecture regarding the subring generated by divisors and the Franchetta conjecture regarding generically defined cycles. As motivation, we show that this Beauville-Voisin-Franchetta conjecture for a hyper-K\"ahler variety $X$ follows from a combination of Grothendieck's standard conjectures for a very general deformation of $X$, Murre's conjecture (D) for $X$ and the Franchetta conjecture for $X^3$. As evidence, beyond the case of Fano varieties of lines on smooth cubic fourfolds, we show that this conjecture holds for codimension-2 and codimension-8 cycles on Lehn-Lehn-Sorger-van Straten eightfolds. Moreover, we establish that the subring of the Chow ring generated by primitive divisors injects into cohomology., Comment: 23 pages, comments very welcome

The integral quality monitor (IQM) system compares the signal measured with a large volume chamber mounted to the linear accelerator's head to the signal calculated using the patient DICOM RT plan for patient-specific quality assurance (PSQA). A method was developed to reconstruct the dose in patients using the signal measured by IQM chamber and patient planning CT images. A software tool named IQMDose3D was implemented to automate this procedure and integrated into the IQM-based PSQA workflow. IQMDose3D enables the physicists to evaluate PSQA by focusing on the clinical perspective by comparing the delivered plan to the approved clinical plan in terms of the clinical goals, dose-volume histogram (DVH) in addition to the three-dimensional (3D) gamma map and gamma pass rate., Comment: Accepted by ICCR 2024 conference

Open cavities are often an essential component in the design of ultra-thin subwavelength metasurfaces and a typical requirement is that cavities have precise, often low frequency, resonances whilst simultaneously being physically compact. To aid this design challenge we develop a methodology to allow isospectral twinning of reference cavities with either smaller or larger ones, enforcing their spectra to coincide so that open resonators are identical in terms of their complex eigenfrequencies. For open systems the spectrum is not purely discrete and real, and we pay special attention to the accurate twinning of leaky modes associated with complex valued eigenfrequencies with an imaginary part orders of magnitude lower than the real part. We further consider twinning of 2D gratings, and model these with Floquet-Bloch conditions along one direction and perfectly matched layers in the other one; complex eigenfrequencies of special interest are located in the vicinity of the positive real line and further depend upon the Bloch wavenumber. The isospectral behaviour is illustrated, and quantified, throughout by numerical simulation using finite element analysis.

While the splinter property is a local property for Noetherian schemes in characteristic zero, Bhatt observed that it imposes strong conditions on the global geometry of proper schemes in positive characteristic. We show that if a proper scheme over a field of positive characteristic is a splinter, then its Nori fundamental group is trivial and its Kodaira dimension is negative. In another direction, Bhatt also showed that any splinter in positive characteristic is a derived splinter. We ask whether the splinter property is a derived-invariant for projective varieties in positive characteristic and give a positive answer for Gorenstein projective varieties with big anticanonical divisor. For that purpose, we introduce the notion of O-equivalence and show that the derived splinter property for schemes of finite type and separated over a fixed Noetherian base is preserved under O-equivalence. Finally, we show that global F-regularity is a derived-invariant for normal Gorenstein projective varieties in positive characteristic., Comment: 35 pages; removed the pseudo-rationality assumption in Theorems F and G. The derived invariance of global F-regularity now holds for normal Gorenstein projective varieties

High-frequency homogenization is used to study dispersive media, containing inclusions placed periodically, for which the properties of the material depend on the frequency (Lorentz or Drude model with damping, for example). Effective properties are obtained near a given point of the dispersion diagram in frequency-wavenumber space. The asymptotic approximations of the dispersion diagrams, and the wavefields, so obtained are then cross-validated via detailed comparison with finite element method simulations in both one and two dimensions.

The study of collaborative multi-agent bandits has attracted significant attention recently. In light of this, we initiate the study of a new collaborative setting, consisting of $N$ agents such that each agent is learning one of $M$ stochastic multi-armed bandits to minimize their group cumulative regret. We develop decentralized algorithms which facilitate collaboration between the agents under two scenarios. We characterize the performance of these algorithms by deriving the per agent cumulative regret and group regret upper bounds. We also prove lower bounds for the group regret in this setting, which demonstrates the near-optimal behavior of the proposed algorithms., Comment: To appear in the proceedings of ICML 2023

Atmospheric composition measurements taken at many high-altitude stations around the world, aim to collect data representative of the free troposphere and of an intercontinental scale. However, the high-altitude environment favours vertical mixing and the transportation of air masses at local or regional scales, which has a potential influence on the composition of the sampled air masses. Mixing processes, source-receptor pathways, and atmospheric chemistry may strongly depend on local and regional weather regimes, and these should be characterized specifically for each station. The Pic du Midi (PDM) isa mountaintop observatory (2850 m a.s.l.) on the north side of the Pyrenees. PDM is associated with the Centre de Recherches Atmosph{\'e}riques (CRA), a site in the foothills ar 600 m a.s.l. 28 km north-east of the PDM. The two centers make up the Pyrenean Platform for the Observation of the Atmosphere (P2OA). Data measured at PDM and CRA were combined to form a5-year hourly dataset of 23 meteorological variables notably: temperature, humidity, cloud cover, wind at several altitudes. The dataset was classified using hierarchical clustering, with the aim of grouping together the days which had similar meteorological characteristics. To complete the clustering, we computed several diagnostic tools, in order to provide additional information and study specific phenomena (foehn, precipitation, atmospheric vertical structure, and thermally driven circulations). This classification resulted in six clusters: three highly populated clusters which correspond to the most frequent meteorological conditions (fair weather, mixed weather and disturbed weather, respectively); a small cluster evidencing clear characteristics of winter northwesterly windstorms; and two small clusters characteristic of south foehn (south- to southwesterly large-scaleflow, associated with warm and dry downslope flow on the lee side of the chain). The diagnostic tools applied to the six clusters provided results in line with the conclusions tentatively drawn from 23 meteorological variables. This, to some extent,validates the approach of hierarchical clustering of local data to distinguish weather regimes. Then statistics of atmospheric composition at PDM were analysed and discussed for each cluster. Radon measurements, notably, revealed that the regional background in the lower troposphere dominates the influence of diurnal thermal flows when daily averaged concentrations are considered. Differences between clusters were demonstrated by the anomalies of CO, CO$_2$, CH$_4$, O$_3$ and aerosol number concentration, and interpretations in relation with chemical sinks and sources are proposed., Comment: Atmospheric Chemistry and Physics, In press

We study point-like explosive events (EE), characterized by emission in the far wings of spectral lines, in a quiet region near the South Pole, using Interface Region Imaging Spectrograph (IRIS) spectra at two slit positions, slit-jaw (SJ) observations and Atmospheric Imaging Assembly (AIA) images. The events were best visible in SiIV spectra; they were weak in SJs, occasionally visible in 1600 A and 304 A AIA images, and invisible in higher temperature AIA images. We identified EEs from position--time images in the far wings of the SiIV lines and measured their distance from the limb. A Gaussian model of the height distribution showed that EEs occur in a narrow (0.9") height range, centered at 3.2" above the continuum limb at 2832.0 A. On the disk, we found that they occur in network boundaries. Further, we studied the line profiles of two bright EEs above the limb and one on the disk. We found that what appears as broad-band emission, is actually a superposition of 2--3 narrow-band Gaussian components with well-separated line profiles, indicating that material is expelled towards and/or away from the observer in discrete episodes in time and in space. The expelled plasma accelerates quickly, reaching line-of-sight (LOS) velocities up to 90 km/s. Overall, the motion was practically along the LOS, as the velocity on the plane of sky was small. In some cases tilted spectra were observed that could be interpreted in terms of rotating motions of up to 30 km/s. We did not find any strong absorption features in the wing of the SiIV lines, although in one case a very weak absorption feature was detected. No motions, indicative of jets, were detected in SJ or AIA images. Reconnection in an asymmetric magnetic-field geometry, in the middle or near the top of small loops is a plausible explanation of their observational characteristics., Comment: Accepted, Solar Physics

An offline OpenPGP user might want to forward part or all of their email messages to third parties. Given that messages are encrypted, this requires transforming them into ciphertexts decryptable by the intended forwarded parties, while maintaining confidentiality and authentication. It is shown in recent lines of work that this can be achieved by means of proxy-re-encryption schemes, however, while encrypted email forwarding is the most mentioned application of proxy-re-encryption, it has not been implemented in the OpenPGP context, to the best of our knowledge. In this paper, we adapt the seminal technique introduced by Blaze, Bleumer and Strauss in EUROCRYPT'98, allowing a Mail Transfer Agent to transform and forward OpenPGP messages without access to decryption keys or plaintexts. We also provide implementation details and a security analysis., Comment: 12 pages, presented at ICMC 2021

Albanese varieties provide a standard tool in algebraic geometry for converting questions about varieties in general, to questions about Abelian varieties. A result of Serre provides the existence of an Albanese variety for any geometrically connected and geometrically reduced scheme of finite type over a field, and a result of Grothendieck--Conrad establishes that Albanese varieties are stable under base change of field provided the scheme is, in addition, proper. A result of Raynaud shows that base change can fail for Albanese varieties without this properness hypothesis. In this paper we show that Albanese varieties of geometrically connected and geometrically reduced schemes of finite type over a field are stable under separable field extensions. We also show that the failure of base change in general is explained by the L/K-image for purely inseparable extensions L/K., Comment: 27 pages, to appear in Annales de l'Institut Fourier