324 results on '"Aubry, A."'
Search Results
2. Differential uniformity of polynomials of degree 10
- Author
-
Aubry, Yves
- Subjects
Mathematics - Number Theory - Abstract
We prove that polynomials of degree 10 over finite fields of even characteristic with some conditions on theirs coefficients have a differential uniformity greater than or equal to 6 over $\mathbb{F}_{2^n}$ for all $n$ sufficiently large.
- Published
- 2024
3. Renormalisation in maximally symmetric spaces and semiclassical gravity in Anti-de Sitter spacetime
- Author
-
Juárez-Aubry, Benito A. and Mamani-Leqque, Milton C.
- Subjects
General Relativity and Quantum Cosmology ,High Energy Physics - Theory - Abstract
We obtain semiclassical gravity solutions in the Poincar\'e fundamental domain of $(3+1)$-dimensional Anti-de Sitter spacetime, PAdS$_4$, with a (massive or massless) Klein-Gordon field (with possibly non-trivial curvature coupling) with Dirichlet or Neumann boundary. Some results are explicitly and graphically presented for special values of the mass and curvature coupling (e.g. minimal or conformal coupling). In order to achieve this, we study in some generality how to perform the Hadamard renormalisation procedure for non-linear observables in maximally symmetric spacetimes in arbitrary dimensions, with emphasis on the stress-energy tensor. We show that, in this maximally symmetric setting, the Hadamard bi-distribution is invariant under the isometries of the spacetime, and can be seen as a `single-argument' distribution depending only on the geodesic distance, which significantly simplifies the Hadamard recursion relations and renormalisation computations., Comment: 16 pages, 2 figures
- Published
- 2024
4. Ultrasound matrix imaging for transcranial in-vivo localization microscopy
- Author
-
Bureau, Flavien, Denis, Louise, Coudert, Antoine, Fink, Mathias, Couture, Olivier, and Aubry, Alexandre
- Subjects
Physics - Medical Physics ,Electrical Engineering and Systems Science - Image and Video Processing ,Physics - Applied Physics - Abstract
Transcranial ultrasound imaging is usually limited by skull-induced attenuation and high-order aberrations. By using contrast agents such as microbubbles in combination with ultrafast imaging, not only can the signal-to-noise ratio be improved, but super-resolution images down to the micrometer scale of the brain vessels can be obtained. However, ultrasound localization microscopy (ULM) remains impacted by wave-front distortions that limit the microbubble detection rate and hamper their localization. In this work, we show how matrix imaging, which relies on the prior recording of the reflection matrix, can provide a solution to those fundamental issues. As an experimental proof-of-concept, an in-vivo reconstruction of deep brain microvessels is performed on three anesthetized sheeps. The compensation of wave distortions is shown to drastically enhance the contrast and resolution of ULM. This experimental study thus opens up promising perspectives for a transcranial and non-ionizing observation of human cerebral microvascular pathologies, such as stroke., Comment: 43 pages, 11 figures, 3 tables
- Published
- 2024
5. Optical matrix imaging applied to embryology
- Author
-
Barolle, Victor, Bureau, Flavien, Guigui, Nicolas, Balondrade, Paul, Brochard, Vincent, Dubois, Olivier, Jouneau, Alice, Bonnet-Garnier, Amélie, and Aubry, Alexandre
- Subjects
Physics - Optics ,Electrical Engineering and Systems Science - Image and Video Processing ,Physics - Medical Physics - Abstract
High-resolution label-free imaging of oocytes and embryos is essential for in vitro fertilization procedures. Yet conventional microscopy fails in this task because of aberrations and multiple scattering induced by refractive index heterogeneities inside the sample. These detrimental phenomena drastically degrade the images of early embryos particularly in depth. To overcome these fundamental problems without sacrificing the frame rate, optical matrix imaging (OMI) is a suitable tool. Relying on an ultra-fast measurement of the reflection matrix associated with the sample, it can compensate for aberration and forward multiple scattering in post-processing, thereby providing three-dimensional and highly contrasted images of embryos at a confocal resolution. As a first proof-of-concept, bovine oocytes and embryos are imaged at a 300 nm resolution almost in real time. Our system enables visualization of intracellular structures such as lipids and mitochondria in the cytoplasm or the zona pellucida surrounding it. Altogether, we demonstrate that OMI is a promising tool for research in developmental biology and for time-lapse monitoring of oocytes and embryos in assisted reproduction., Comment: 18 pages, 6 figures
- Published
- 2024
6. General Detection-based Text Line Recognition
- Author
-
Baena, Raphael, Kalleli, Syrine, and Aubry, Mathieu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
We introduce a general detection-based approach to text line recognition, be it printed (OCR) or handwritten (HTR), with Latin, Chinese, or ciphered characters. Detection-based approaches have until now been largely discarded for HTR because reading characters separately is often challenging, and character-level annotation is difficult and expensive. We overcome these challenges thanks to three main insights: (i) synthetic pre-training with sufficiently diverse data enables learning reasonable character localization for any script; (ii) modern transformer-based detectors can jointly detect a large number of instances, and, if trained with an adequate masking strategy, leverage consistency between the different detections; (iii) once a pre-trained detection model with approximate character localization is available, it is possible to fine-tune it with line-level annotation on real data, even with a different alphabet. Our approach, dubbed DTLR, builds on a completely different paradigm than state-of-the-art HTR methods, which rely on autoregressive decoding, predicting character values one by one, while we treat a complete line in parallel. Remarkably, we demonstrate good performance on a large range of scripts, usually tackled with specialized approaches. In particular, we improve state-of-the-art performances for Chinese script recognition on the CASIA v2 dataset, and for cipher recognition on the Borg and Copiale datasets. Our code and models are available at https://github.com/raphael-baena/DTLR.
- Published
- 2024
7. Reflection Matrix Imaging for Wave Velocity Tomography
- Author
-
Bureau, Flavien, Giraudat, Elsa, Ber, Arthur Le, Lambert, William, Carmier, Louis, Guibal, Aymeric, Fink, Mathias, and Aubry, Alexandre
- Subjects
Physics - Medical Physics ,Electrical Engineering and Systems Science - Image and Video Processing ,Physics - Applied Physics - Abstract
Besides controlling wave trajectory inside complex media, wave velocity constitutes a relevant bio-marker for medical imaging. In a transmission configuration, wave-front distortions can be unscrambled to provide a map of the wave velocity landscape $c(\mathbf{r})$. However, most in-vivo applications correspond to a reflection configuration for which only back-scattered echoes generated by short-scale fluctuations of $c(\mathbf{r})$ can be harvested. Under a single scattering assumption, this speckle wave-field cannot provide the long-scale variations of $c(\mathbf{r})$. In this paper, we go beyond the first Born approximation and show how a map of $c(\mathbf{r})$ can be retrieved in epi-detection. To that aim, a reflection matrix approach of wave imaging is adopted. While standard reflection imaging methods generally rely on confocal focusing operations, matrix imaging consists in decoupling the location of the incident and received focal spots. Following this principle, a self-portrait of the focusing process can be obtained around each point of the medium. The Gouy phase shift exhibited by each focal spot is leveraged to finely monitor the wave velocity distribution $c(\mathbf{r})$ inside the medium. Experiment in a tissue-mimicking phantom and numerical simulations are first presented to validate our method. Speed-of-sound tomography is then applied to ultrasound data collected on the liver of a difficult-to-image patient. Beyond paving the way towards quantitative ultrasound, our approach can also be extremely rewarding for standard imaging. Indeed, each echo can be assigned to the actual position of a scatterer. It allows an absolute measurement of distance, an observable often used for diagnosis but generally extremely sensitive to wave velocity fluctuations., Comment: 45 pages, 9 figures, 3 tables
- Published
- 2024
8. Detecting Looted Archaeological Sites from Satellite Image Time Series
- Author
-
Vincent, Elliot, Saroufim, Mehraïl, Chemla, Jonathan, Ubelmann, Yves, Marquis, Philippe, Ponce, Jean, and Aubry, Mathieu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Archaeological sites are the physical remains of past human activity and one of the main sources of information about past societies and cultures. However, they are also the target of malevolent human actions, especially in countries having experienced inner turmoil and conflicts. Because monitoring these sites from space is a key step towards their preservation, we introduce the DAFA Looted Sites dataset, \datasetname, a labeled multi-temporal remote sensing dataset containing 55,480 images acquired monthly over 8 years across 675 Afghan archaeological sites, including 135 sites looted during the acquisition period. \datasetname~is particularly challenging because of the limited number of training samples, the class imbalance, the weak binary annotations only available at the level of the time series, and the subtlety of relevant changes coupled with important irrelevant ones over a long time period. It is also an interesting playground to assess the performance of satellite image time series (SITS) classification methods on a real and important use case. We evaluate a large set of baselines, outline the substantial benefits of using foundation models and show the additional boost that can be provided by using complete time series instead of using a single image.
- Published
- 2024
9. An Interpretable Deep Learning Approach for Morphological Script Type Analysis
- Author
-
Vlachou-Efstathiou, Malamatenia, Siglidis, Ioannis, Stutzmann, Dominique, and Aubry, Mathieu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Defining script types and establishing classification criteria for medieval handwriting is a central aspect of palaeographical analysis. However, existing typologies often encounter methodological challenges, such as descriptive limitations and subjective criteria. We propose an interpretable deep learning-based approach to morphological script type analysis, which enables systematic and objective analysis and contributes to bridging the gap between qualitative observations and quantitative measurements. More precisely, we adapt a deep instance segmentation method to learn comparable character prototypes, representative of letter morphology, and provide qualitative and quantitative tools for their comparison and analysis. We demonstrate our approach by applying it to the Textualis Formata script type and its two subtypes formalized by A. Derolez: Northern and Southern Textualis, Comment: Accepted at ICDAR 2024 Workshop on Computational Paleography (IWCP, 31 August - Athens, Greece)
- Published
- 2024
10. Magazine Supply Optimization: a Case-study
- Author
-
Nguyen, Duong, Ulianovici, Ana, Achour, Sami, Aubry, Soline, and Chesneau, Nicolas
- Subjects
Computer Science - Artificial Intelligence ,Mathematics - Optimization and Control - Abstract
Supply optimization is a complex and challenging task in the magazine retail industry because of the fixed inventory assumption, irregular sales patterns, and varying product and point-of-sale characteristics. We introduce AthenIA, an industrialized magazine supply optimization solution that plans the supply for over 20,000 points of sale in France. We modularize the supply planning process into a four-step pipeline: demand sensing, optimization, business rules, and operating. The core of the solution is a novel group conformalized quantile regression method that integrates domain expert insights, coupled with a supply optimization technique that balances the costs of out-of-stock against the costs of over-supply. AthenIA has proven to be a valuable tool for magazine publishers, particularly in the context of evolving economic and ecological challenges.
- Published
- 2024
11. Historical Printed Ornaments: Dataset and Tasks
- Author
-
Chaki, Sayan Kumar, Baltaci, Zeynep Sonat, Vincent, Elliot, Emonet, Remi, Vial-Bonacci, Fabienne, Bahier-Porte, Christelle, Aubry, Mathieu, and Fournel, Thierry
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
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/.
- Published
- 2024
12. Targeted energy transfer dynamics and chemical reactions
- Author
-
Almazova, N., Aubry, S., and Tsironis, G. P.
- Subjects
Physics - Chemical Physics - Abstract
Ultrafast reaction processes take place when resonant features of nonlinear model systems are taken into account. In the targeted energy or electron transfer dimer model this is accomplished through the implementation of nonlinear oscillators with opposing types of nonlinearities, one attractive while the second repulsive. In the present work we show that this resonant behavior survives if we take into account the vibrational degrees of freedom as well. After giving a summary on the basic formalism of chemical reactions we show that resonant electron transfer can be assisted by vibrations. We find the condition for this efficient transfer and show that in the case of additional interaction with noise a distinct non-Arrhenius behavior develops that is markedly different from the usual Kramers-like activated transfer.
- Published
- 2024
13. Diffusion Models as Data Mining Tools
- Author
-
Siglidis, Ioannis, Holynski, Aleksander, Efros, Alexei A., Aubry, Mathieu, and Ginosar, Shiry
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Artificial Intelligence - Abstract
This paper demonstrates how to use generative models trained for image synthesis as tools for visual data mining. Our insight is that since contemporary generative models learn an accurate representation of their training data, we can use them to summarize the data by mining for visual patterns. Concretely, we show that after finetuning conditional diffusion models to synthesize images from a specific dataset, we can use these models to define a typicality measure on that dataset. This measure assesses how typical visual elements are for different data labels, such as geographic location, time stamps, semantic labels, or even the presence of a disease. This analysis-by-synthesis approach to data mining has two key advantages. First, it scales much better than traditional correspondence-based approaches since it does not require explicitly comparing all pairs of visual elements. Second, while most previous works on visual data mining focus on a single dataset, our approach works on diverse datasets in terms of content and scale, including a historical car dataset, a historical face dataset, a large worldwide street-view dataset, and an even larger scene dataset. Furthermore, our approach allows for translating visual elements across class labels and analyzing consistent changes., Comment: Project Page: https://diff-mining.github.io/ Accepted in ECCV 2024
- Published
- 2024
14. Transformer Block Coupling and its Correlation with Generalization in LLMs
- Author
-
Aubry, Murdock, Meng, Haoming, Sugolov, Anton, and Papyan, Vardan
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence ,Computer Science - Computation and Language - Abstract
Large Language Models (LLMs) have made significant strides in natural language processing, and a precise understanding of the internal mechanisms driving their success is essential. In this work, we trace the trajectories of individual tokens as they pass through transformer blocks, and linearize the system along these trajectories through their Jacobian matrices. By examining the relationships between these Jacobians, we uncover a $\textbf{transformer block coupling}$ phenomenon in a variety of LLMs, characterized by the coupling of their top singular vectors across tokens and depth. Our findings reveal that coupling $\textit{positively correlates}$ with model performance, and that this relationship is stronger than with other hyperparameters, namely parameter budget, model depth, and embedding dimension. We further investigate the emergence of these properties through training, noting the development of coupling, as well as an increase in linearity and layer-wise exponential growth in the token trajectories. These collective insights provide a novel perspective on the interactions between token embeddings, and prompt further approaches to study training and generalization in LLMs.
- Published
- 2024
15. Satellite Image Time Series Semantic Change Detection: Novel Architecture and Analysis of Domain Shift
- Author
-
Vincent, Elliot, Ponce, Jean, and Aubry, Mathieu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Satellite imagery plays a crucial role in monitoring changes happening on Earth's surface and aiding in climate analysis, ecosystem assessment, and disaster response. In this paper, we tackle semantic change detection with satellite image time series (SITS-SCD) which encompasses both change detection and semantic segmentation tasks. We propose a new architecture that improves over the state of the art, scales better with the number of parameters, and leverages long-term temporal information. However, for practical use cases, models need to adapt to spatial and temporal shifts, which remains a challenge. We investigate the impact of temporal and spatial shifts separately on global, multi-year SITS datasets using DynamicEarthNet and MUDS. We show that the spatial domain shift represents the most complex setting and that the impact of temporal shift on performance is more pronounced on change detection than on semantic segmentation, highlighting that it is a specific issue deserving further attention.
- Published
- 2024
16. The Hadamard condition on a Cauchy surface and the renormalized stress-energy tensor
- Author
-
Juárez-Aubry, Benito A., Kay, Bernard S., Miramontes, Tonatiuh, and Sudarsky, Daniel
- Subjects
General Relativity and Quantum Cosmology ,High Energy Physics - Theory ,Mathematical Physics - Abstract
Given a Cauchy surface in a curved spacetime and a suitably defined quantum state on the CCR algebra of the Klein-Gordon quantum field on that surface, we show, by expanding the squared spacetime geodesic distance and the `$U$' and `$V$' Hadamard coefficients (and suitable derivatives thereof) in sufficiently accurate covariant Taylor expansions on the surface that the renormalized expectation value of the quantum stress-energy tensor on the surface is determined by the geometry of the surface and the first 4 time derivatives of the metric off the surface, in addition to the Cauchy data for the field's two-point function. This result has been anticipated in and is motivated by a previous investigation by the authors on the initial value problem in semiclassical gravity, for which the geometric initial data corresponds {\it a priori} to the metric on the surface and up to 3 time derivatives off the surface, but where it was argued that the fourth derivative can be obtained with aid of the field equations on the initial surface., Comment: 65 pages, discussion improved
- Published
- 2024
- Full Text
- View/download PDF
17. BEAST DB: Grand-Canonical Database of Electrocatalyst Properties
- Author
-
Tezak, Cooper, Clary, Jacob, Gerits, Sophie, Quinton, Joshua, Rich, Benjamin, Singstock, Nicholas, Alherz, Abdulaziz, Aubry, Taylor, Clark, Struan, Hurst, Rachel, Del Ben, Mauro, Sutton, Christopher, Sundararaman, Ravishankar, Musgrave, Charles, and Vigil-Fowler, Derek
- Subjects
Condensed Matter - Materials Science ,Physics - Chemical Physics - Abstract
We present BEAST DB, an open-source database comprised of ab initio electrochemical data computed using grand-canonical density functional theory in implicit solvent at consistent calculation parameters. The database contains over 20,000 surface calculations and covers a broad set of heterogeneous catalyst materials and electrochemical reactions. Calculations were performed at self-consistent fixed potential as well as constant charge to facilitate comparisons to the computational hydrogen electrode. This article presents common use cases of the database to rationalize trends in catalyst activity, screen catalyst material spaces, understand elementary mechanistic steps, analyze electronic structure, and train machine learning models to predict higher fidelity properties. Users can interact graphically with the database by querying for individual calculations to gain granular understanding of reaction steps or by querying for an entire reaction pathway on a given material using an interactive reaction pathway tool. BEAST DB will be periodically updated, with planned future updates to include advanced electronic structure data, surface speciation studies, and greater reaction coverage., Comment: 24 pages, 8 figures
- Published
- 2024
18. Trinomials with high differential uniformity
- Author
-
Aubry, Yves, Herbaut, Fabien, and Issa, Ali
- Subjects
Mathematics - Number Theory - Abstract
The context of this work is the study of the differential uniformity of polynomials defined over finite fields of even characteristic. We provide infinite families of trinomials with high differential uniformity when the base field is large enough. It means in particular that these trinomials are not exceptional almost perfect nonlinear.
- Published
- 2024
19. Historical Astronomical Diagrams Decomposition in Geometric Primitives
- Author
-
Kalleli, Syrine, Trigg, Scott, Albouy, Ségolène, Husson, Mathieu, and Aubry, Mathieu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Automatically extracting the geometric content from the hundreds of thousands of diagrams drawn in historical manuscripts would enable historians to study the diffusion of astronomical knowledge on a global scale. However, state-of-the-art vectorization methods, often designed to tackle modern data, are not adapted to the complexity and diversity of historical astronomical diagrams. Our contribution is thus twofold. First, we introduce a unique dataset of 303 astronomical diagrams from diverse traditions, ranging from the XIIth to the XVIIIth century, annotated with more than 3000 line segments, circles and arcs. Second, we develop a model that builds on DINO-DETR to enable the prediction of multiple geometric primitives. We show that it can be trained solely on synthetic data and accurately predict primitives on our challenging dataset. Our approach widely improves over the LETR baseline, which is restricted to lines, by introducing a meaningful parametrization for multiple primitives, jointly training for detection and parameter refinement, using deformable attention and training on rich synthetic data. Our dataset and code are available on our webpage., Comment: Code and dataset are available in http://imagine.enpc.fr/~kallelis/icdar2024/
- Published
- 2024
20. ITRUSST Consensus on Standardised Reporting for Transcranial Ultrasound Stimulation
- Author
-
Martin, Eleanor, Aubry, Jean-François, Schafer, Mark, Verhagen, Lennart, Treeby, Bradley, and Pauly, Kim Butts
- Subjects
Physics - Medical Physics ,Physics - Biological Physics - Abstract
As transcranial ultrasound stimulation (TUS) advances as a precise, non-invasive neuromodulatory method, there is a need for consistent reporting standards to enable comparison and reproducibility across studies. To this end, the International Transcranial Ultrasonic Stimulation Safety and Standards Consortium (ITRUSST) formed a subcommittee of experts across several domains to review and suggest standardised reporting parameters for low intensity TUS, resulting in the guide presented here. The scope of the guide is limited to reporting the ultrasound aspects of a study. The guide and supplementary material provide a simple checklist covering the reporting of: (1) the transducer and drive system, (2) the drive system settings, (3) the free field acoustic parameters, (4) the pulse timing parameters, (5) \emph{in situ} estimates of exposure parameters in the brain, and (6) intensity parameters. Detailed explanations for each of the parameters, including discussions on assumptions, measurements, and calculations, are also provided., Comment: 23 pages, 4 figures, 4 tables, stand alone checklist
- Published
- 2024
21. Maximum number of rational points on hypersurfaces in weighted projective spaces over finite fields
- Author
-
Aubry, Yves and Perret, Marc
- Subjects
Mathematics - Algebraic Geometry - Abstract
An upper bound for the maximum number of rational points on an hypersurface in a projective space over a finite field has been conjectured by Tsfasman and proved by Serre in 1989. The analogue question for hypersurfaces on weighted projective spaces has been considered by Castryck, Ghorpade, Lachaud, O'Sullivan, Ram and the first author in 2017. A conjecture has been proposed there and proved in the particular case of the dimension 2. We prove here the conjecture in any dimension provided the second weight is also equal to one.
- Published
- 2024
22. Reflection Measurement of the Scattering Mean Free Path at the Onset of Multiple Scattering
- Author
-
Goïcoechea, Antton, Brütt, Cécile, Ber, Arthur Le, Bureau, Flavien, Lambert, William, Prada, Claire, Derode, Arnaud, and Aubry, Alexandre
- Subjects
Physics - Medical Physics ,Condensed Matter - Disordered Systems and Neural Networks ,Physics - Applied Physics - Abstract
Multiple scattering of waves presents challenges for imaging complex media but offers potential for their characterization. Its onset is actually governed by the scattering mean free path $\ell_s$ that provides crucial information on the medium micro-architecture. Here, we introduce a reflection matrix method designed to estimate this parameter from the time decay of the single scattering rate. Our method is first validated by an ultrasound experiment on a tissue-mimicking phantom before being applied in-vivo to a human liver. This study opens important perspectives for quantitative imaging of heterogeneous media with waves, whether it be for non-destructive testing, biomedical or geophysical applications., Comment: 29 pages, 7 figures
- Published
- 2024
- Full Text
- View/download PDF
23. FocalPose++: Focal Length and Object Pose Estimation via Render and Compare
- Author
-
Cífka, Martin, Ponimatkin, Georgy, Labbé, Yann, Russell, Bryan, Aubry, Mathieu, Petrik, Vladimir, and Sivic, Josef
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
We introduce FocalPose++, a neural render-and-compare method for jointly estimating the camera-object 6D pose and camera focal length given a single RGB input image depicting a known object. The contributions of this work are threefold. First, we derive a focal length update rule that extends an existing state-of-the-art render-and-compare 6D pose estimator to address the joint estimation task. Second, we investigate several different loss functions for jointly estimating the object pose and focal length. We find that a combination of direct focal length regression with a reprojection loss disentangling the contribution of translation, rotation, and focal length leads to improved results. Third, we explore the effect of different synthetic training data on the performance of our method. Specifically, we investigate different distributions used for sampling object's 6D pose and camera's focal length when rendering the synthetic images, and show that parametric distribution fitted on real training data works the best. We show results on three challenging benchmark datasets that depict known 3D models in uncontrolled settings. We demonstrate that our focal length and 6D pose estimates have lower error than the existing state-of-the-art methods., Comment: 25 pages, 22 figures. IEEE TPAMI, 2024. Extended version of the conference paper arXiv:2204.05145
- Published
- 2023
- Full Text
- View/download PDF
24. A solver for linear scalar ordinary differential equations whose running time is bounded independent of frequency
- Author
-
Aubry, Murdock and Bremer, James
- Subjects
Mathematics - Numerical Analysis - Abstract
When the eigenvalues of the coefficient matrix for a linear scalar ordinary differential equation are of large magnitude, its solutions exhibit complicated behaviour, such as high-frequency oscillations, rapid growth or rapid decay. The cost of representing such solutions using standard techniques grows with the magnitudes of the eigenvalues. As a consequence, the running times of most solvers for ordinary differential equations also grow with these eigenvalues. However, a large class of scalar ordinary differential equations with slowly-varying coefficients admit slowly-varying phase functions that can be represented at a cost which is bounded independent of the magnitudes of the eigenvalues of the corresponding coefficient matrix. Here, we introduce a numerical algorithm for constructing slowly-varying phase functions which represent the solutions of a linear scalar ordinary differential equation. Our method's running time depends on the complexity of the equation's coefficients, but is bounded independent of the magnitudes of the equation's eigenvalues. Once the phase functions have been constructed, essentially any reasonable initial or boundary value problem for the scalar equation can be easily solved. We present the results of numerical experiments showing that, despite its greater generality, our algorithm is competitive with state-of-the-art methods for solving highly-oscillatory second order differential equations. We also compare our method with Magnus-type exponential integrators and find that our approach is orders of magnitude faster in the high-frequency regime., Comment: arXiv admin note: text overlap with arXiv:2308.03288, arXiv:2309.13848
- Published
- 2023
25. ITRUSST Consensus on Biophysical Safety for Transcranial Ultrasonic Stimulation
- Author
-
Aubry, Jean-Francois, Attali, David, Schafer, Mark, Fouragnan, Elsa, Caskey, Charles, Chen, Robert, Darmani, Ghazaleh, Bubrick, Ellen J., Sallet, Jérôme, Butler, Christopher, Stagg, Charlotte, Klein-Flügge, Miriam, Yoo, Seung-Schik, Treeby, Brad, Verhagen, Lennart, and Pauly, Kim Butts
- Subjects
Physics - Biological Physics ,Physics - Medical Physics - Abstract
Transcranial ultrasonic stimulation (TUS) is an emerging technology for non-invasive brain stimulation. In a series of meetings, the International Consortium for Transcranial Ultrasonic Stimulation Safety and Standards (ITRUSST) has established expert consensus on considerations for the biophysical safety of TUS, drawing upon the relevant diagnostic ultrasound literature and regulations. This report reflects a consensus expert opinion and can inform but not replace regulatory guidelines or official international standards. Their establishment by international and national commissions will follow expert consensus. Similarly, this consensus will inform but not replace ethical evaluation, which will consider aspects beyond biophysical safety relevant to burden, risk, and benefit, such as physiological effects and disease-specific interactions. Here, we assume the application of TUS to persons who are not at risk for thermal or mechanical damage, and without ultrasound contrast agents. In this context, we present a concise yet comprehensive set of levels for a nonsignificant risk of TUS application. For mechanical effects, it is safe if the mechanical index (MI) or the mechanical index for transcranial application (MItc) does not exceed 1.9. For thermal effects, it is safe if any of the following three levels are met: a temperature rise less than 2 C, a thermal dose less than 0.25 CEM43, or specific values of the thermal index (TI) for a given exposure time. We review literature relevant to our considerations and discuss limitations and future developments of our approach., Comment: ITRUSST consensus, 15 pages, 1 table, 1 figure
- Published
- 2023
26. Unveiling the deep plumbing system of a volcano by a reflection matrix analysis of seismic noise
- Author
-
Giraudat, Elsa, Burtin, Arnaud, Ber, Arthur Le, Fink, Mathias, Komorowski, Jean-Christophe, and Aubry, Alexandre
- Subjects
Physics - Geophysics ,Electrical Engineering and Systems Science - Image and Video Processing ,Physics - Applied Physics - Abstract
In geophysics, volcanoes are particularly difficult to image because of the multi-scale heterogeneities of fluids and rocks that compose them and their complex non-linear dynamics. By exploiting seismic noise recorded by a sparse array of geophones, we are able to reveal the magmatic and hydrothermal plumbing system of La Soufri\`ere volcano in Guadeloupe. Spatio-temporal cross-correlation of seismic noise actually provides the impulse responses between virtual geophones located inside the volcano. The resulting reflection matrix can be exploited to numerically perform an auto-focus of seismic waves on any reflector of the underground. An unprecedented view on the volcano's inner structure is obtained at a half-wavelength resolution. This innovative observable provides fundamental information for the conceptual modeling and high-resolution monitoring of volcanoes., Comment: 32 pages, 9 figures
- Published
- 2023
27. Closed points on curves over finite fields
- Author
-
Aubry, Yves, Herbaut, Fabien, and Monaldi, Julien
- Subjects
Mathematics - Algebraic Geometry - Abstract
We are interested in the quantity $\rho$(q, g) defined as the smallest positive integer such that r $\ge$ $\rho$(q, g) implies that any absolutely irreducible smooth projective algebraic curve defined over F q of genus g has a closed point of degree r. We provide general upper bounds for this number and its exact value for g = 1, 2 and 3. We also improve the known upper bounds on the number of closed points of degree 2 on a curve.
- Published
- 2023
28. Multi-Spectral Reflection Matrix for Ultra-Fast 3D Label-Free Microscopy
- Author
-
Balondrade, Paul, Barolle, Victor, Guigui, Nicolas, Auriant, Emeric, Rougier, Nathan, Boccara, Claude, Fink, Mathias, and Aubry, Alexandre
- Subjects
Physics - Optics ,Electrical Engineering and Systems Science - Image and Video Processing - Abstract
Label-free microscopy exploits light scattering to obtain a three-dimensional image of biological tissues. However, light propagation is affected by aberrations and multiple scattering, which drastically degrade the image quality and limit the penetration depth. Multi-conjugate adaptive optics and time-gated matrix approaches have been developed to compensate for aberrations but the associated frame rate is extremely limited for 3D imaging. Here we develop a multi-spectral matrix approach to solve these fundamental problems. Based on a sparse illumination scheme and an interferometric measurement of the reflected wave-field at multiple wavelengths, the focusing process can be optimized in post-processing for any voxel by addressing independently each frequency component of the reflection matrix. A proof-of-concept experiment demonstrates the three-dimensional image of an opaque human cornea over a 0.1 mm$^3$-field-of-view at a 290 nm-resolution and a 1 Hz-frame rate. This work paves the way towards a fully-digital microscope allowing real-time, in-vivo, quantitative and deep inspection of tissues., Comment: 55 pages, 9 figures
- Published
- 2023
- Full Text
- View/download PDF
29. 3D Localization and Tracking Methods for Multi-Platform Radar Networks
- Author
-
Marino, Angela, Soldi, Giovanni, Gaglione, Domenico, Aubry, Augusto, Braca, Paolo, De Maio, Antonio, and Willett, Peter
- Subjects
Electrical Engineering and Systems Science - Signal Processing - Abstract
Multi-platform radar networks (MPRNs) are an emerging sensing technology due to their ability to provide improved surveillance capabilities over plain monostatic and bistatic systems. The design of advanced detection, localization, and tracking algorithms for efficient fusion of information obtained through multiple receivers has attracted much attention. However, considerable challenges remain. This article provides an overview on recent unconstrained and constrained localization techniques as well as multitarget tracking (MTT) algorithms tailored to MPRNs. In particular, two data-processing methods are illustrated and explored in detail, one aimed at accomplishing localization tasks the other tracking functions. As to the former, assuming a MPRN with one transmitter and multiple receivers, the angular and range constrained estimator (ARCE) algorithm capitalizes on the knowledge of the transmitter antenna beamwidth. As to the latter, the scalable sum-product algorithm (SPA) based MTT technique is presented. Additionally, a solution to combine ARCE and SPA-based MTT is investigated in order to boost the accuracy of the overall surveillance system. Simulated experiments show the benefit of the combined algorithm in comparison with the conventional baseline SPA-based MTT and the stand-alone ARCE localization, in a 3D sensing scenario.
- Published
- 2023
30. The Levin approach to the numerical calculation of phase functions
- Author
-
Aubry, Murdock and Bremer, James
- Subjects
Mathematics - Numerical Analysis - Abstract
The solutions of scalar ordinary differential equations become more complex as their coefficients increase in magnitude. As a consequence, when a standard solver is applied to such an equation, its running time grows with the magnitudes of the equation's coefficients. It is well known, however, that scalar ordinary differential equations with slowly-varying coefficients admit slowly-varying phase functions whose cost to represent via standard techniques is largely independent of the magnitude of the equation's coefficients. This observation is the basis of most methods for the asymptotic approximation of the solutions of ordinary differential equations, including the WKB method. Here, we introduce two numerical algorithms for constructing phase functions for scalar ordinary differential equations inspired by the classical Levin method for the calculation of oscillatory integrals. In the case of a large class of scalar ordinary differential equations with slowly-varying coefficients, their running times are independent of the magnitude of the equation's coefficients. The results of extensive numerical experiments demonstrating the properties of our algorithms are presented.
- Published
- 2023
31. Strong localization of microwaves beyond 2D in aperiodic Vogel spirals
- Author
-
Razo-López, Luis A., Aubry, Geoffroy J., Pinheiro, Felipe A., and Mortessagne, Fabrice
- Subjects
Condensed Matter - Disordered Systems and Neural Networks ,Physics - Optics - Abstract
We carry out dynamical microwave transport experiments in aperiodic Vogel spiral arrays of cylinders with high dielectric permittivity. We experimentally disclose the electromagnetic modal structure of these structures in real space showing that they simultaneously support long-lived modes with Gaussian, exponential, and power law spatial decay. This unique modal structure, which cannot be found in traditional periodic or disordered photonic materials, is shown to be at the origin of strong localization in Vogel spirals that survives even in three dimensions. Altogether our results unveil the manifestations of the rich, unprecedented, spatial structure of electromagnetic modes supported by aperiodic photonic systems in wave transport and localization., Comment: main text and Supplemental Material: 10 pages and 10 figures
- Published
- 2023
32. Differentiable Blocks World: Qualitative 3D Decomposition by Rendering Primitives
- Author
-
Monnier, Tom, Austin, Jake, Kanazawa, Angjoo, Efros, Alexei A., and Aubry, Mathieu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Given a set of calibrated images of a scene, we present an approach that produces a simple, compact, and actionable 3D world representation by means of 3D primitives. While many approaches focus on recovering high-fidelity 3D scenes, we focus on parsing a scene into mid-level 3D representations made of a small set of textured primitives. Such representations are interpretable, easy to manipulate and suited for physics-based simulations. Moreover, unlike existing primitive decomposition methods that rely on 3D input data, our approach operates directly on images through differentiable rendering. Specifically, we model primitives as textured superquadric meshes and optimize their parameters from scratch with an image rendering loss. We highlight the importance of modeling transparency for each primitive, which is critical for optimization and also enables handling varying numbers of primitives. We show that the resulting textured primitives faithfully reconstruct the input images and accurately model the visible 3D points, while providing amodal shape completions of unseen object regions. We compare our approach to the state of the art on diverse scenes from DTU, and demonstrate its robustness on real-life captures from BlendedMVS and Nerfstudio. We also showcase how our results can be used to effortlessly edit a scene or perform physical simulations. Code and video results are available at https://www.tmonnier.com/DBW ., Comment: Project webpage with code and videos: https://www.tmonnier.com/DBW. V2 update includes comparisons based on NeuS, hyperparameter analysis and failure cases
- Published
- 2023
33. New Methods for MLE of Toeplitz Structured Covariance Matrices with Applications to RADAR Problems
- Author
-
Aubry, Augusto, Babu, Prabhu, De Maio, Antonio, and Rosamilia, Massimo
- Subjects
Electrical Engineering and Systems Science - Signal Processing - Abstract
This work considers Maximum Likelihood Estimation (MLE) of a Toeplitz structured covariance matrix. In this regard, an equivalent reformulation of the MLE problem is introduced and two iterative algorithms are proposed for the optimization of the equivalent statistical learning framework. Both the strategies are based on the Majorization Minimization (MM) paradigm and hence enjoy nice properties such as monotonicity and ensured convergence to a stationary point of the equivalent MLE problem. The proposed framework is also extended to deal with MLE of other practically relevant covariance structures, namely, the banded Toeplitz, block Toeplitz, and Toeplitz-block-Toeplitz. Through numerical simulations, it is shown that the new methods provide excellent performance levels in terms of both mean square estimation error (which is very close to the benchmark Cram\'er-Rao Bound (CRB)) and signal-to-interference-plus-noise ratio, especially in comparison with state of the art strategies., Comment: submitted to IEEE Transactions on Signal Processing. arXiv admin note: substantial text overlap with arXiv:2110.12176
- Published
- 2023
34. Power-Aperture Resource Allocation for a MPAR with Communications Capabilities
- Author
-
Aubry, Augusto, De Maio, Antonio, and Pallotta, Luca
- Subjects
Electrical Engineering and Systems Science - Signal Processing - Abstract
Multifunction phased array radars (MPARs) exploit the intrinsic flexibility of their active electronically steered array (ESA) to perform, at the same time, a multitude of operations, such as search, tracking, fire control, classification, and communications. This paper aims at addressing the MPAR resource allocation so as to satisfy the quality of service (QoS) demanded by both line of sight (LOS) and reflective intelligent surfaces (RIS)-aided non line of sight (NLOS) search operations along with communications tasks. To this end, the ranges at which the cumulative detection probability and the channel capacity per bandwidth reach a desired value are introduced as task quality metrics for the search and communication functions, respectively. Then, to quantify the satisfaction level of each task, for each of them a bespoke utility function is defined to map the associated quality metric into the corresponding perceived utility. Hence, assigning different priority weights to each task, the resource allocation problem, in terms of radar power aperture (PAP) specification, is formulated as a constrained optimization problem whose solution optimizes the global radar QoS. Several simulations are conducted in scenarios of practical interest to prove the effectiveness of the approach., Comment: 14 pages, 16 figures
- Published
- 2023
- Full Text
- View/download PDF
35. A Learning-Inspired Strategy to Design Binary Sequences with Good Correlation Properties: SISO and MIMO Radar Systems
- Author
-
Rezaei, Omid, Ahmadi, Mahdi, Naghsh, Mohammad Mahdi, Aubry, Augusto, Nayebi, Mohammad Mahdi, and De Maio, Antonio
- Subjects
Electrical Engineering and Systems Science - Signal Processing - Abstract
In this paper, the design of binary sequences exhibiting low values of aperiodic/periodic correlation functions, in terms of Integrated Sidelobe Level (ISL), is pursued via a learning-inspired method. Specifcally, the synthesis of either a single or a burst of codes is addressed, with reference to both Single-Input Single-Output (SISO) and Multiple-Input Multiple-Output (MIMO) radar systems. Two optimization machines, referred to as two-layer and single-layer Binary Sequence Correlation Network (BiSCorN), able to learn actions to design binary sequences with small ISL/Complementary ISL (CISL) for SISO and MIMO systems are proposed. These two networks differ in terms of the capability to synthesize Low-Correlation-Zone (LCZ) sequences and computational cost. Numerical experiments show that proposed techniques can outperform state-of-the-art algorithms for the design of binary sequences and Complementary Sets of Sequences (CSS) in terms of ISL and, interestingly, of Peak Sidelobe Level (PSL).
- Published
- 2023
36. A Novel Mechanism for the Formation of Dislocation Cell Patterns in BCC Metal
- Author
-
Cho, Jaehyun, Hsiung, Luke L., Rudd, Robert E., and Aubry, Sylvie
- Subjects
Condensed Matter - Materials Science ,High Energy Physics - Theory - Abstract
In this study, we present the first simulation results of the formation of dislocation cell wall microstructures in tantalum subjected to shock loading. Dislocation patterns and cell wall formation are important to understanding the mechanical properties of the materials in which they spontaneously arise, and yet the processing and self-assembly mechanisms leading to their formation are poorly understood. By employing transmission electron microscopy and discrete dislocation dynamics, we propose a new mechanism involving coplanar dislocations and pseudo-dipole mixed dislocation arrays that is essential to the pattern formation process. Our large-scale 3D DDD simulations demonstrate the self-organization of dislocation networks into cell walls in deformed BCC metal (tantalum) persisting at the strain 20%. The simulation analysis captures several crucial aspects of how the dislocation cell pattern affects metal plasticity, as observed in experiments. Although experimental evidence is inconclusive regarding whether cell wall formation takes place at the shock front, after the shock, during release, or when the sample has had enough time to relax post-recovery, our simulations indicate cell wall formation occurs after the shock and before release. The extended Taylor hardening composite model effectively considers the non-uniform dislocation density when cell walls form and accurately describes the corresponding flow stress.
- Published
- 2023
37. Quantum strong cosmic censorship and black hole evaporation
- Author
-
Juárez-Aubry, Benito A.
- Subjects
General Relativity and Quantum Cosmology ,High Energy Physics - Theory ,Mathematical Physics - Abstract
It is common folklore that semiclassical arguments suggest that in black hole evaporation an initially pure state can become mixed. This is known as the \emph{information loss puzzle} (or {\it paradox}). Here we argue that, if taken at face value, semiclassical gravity suggests the formation of a final singularity instead of information loss. A quantum strong cosmic censorship conjecture, for which we give a rigorous statement, supports this conclusion. Thus, there are no reasons to expect a failure of unitarity in black hole evaporation or for any quantum gravity theory that can `cure' singularities., Comment: 9 pages, 3 figures. Comments are welcome!
- Published
- 2023
- Full Text
- View/download PDF
38. Learnable Earth Parser: Discovering 3D Prototypes in Aerial Scans
- Author
-
Loiseau, Romain, Vincent, Elliot, Aubry, Mathieu, and Landrieu, Loic
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
We propose an unsupervised method for parsing large 3D scans of real-world scenes with easily-interpretable shapes. This work aims to provide a practical tool for analyzing 3D scenes in the context of aerial surveying and mapping, without the need for user annotations. Our approach is based on a probabilistic reconstruction model that decomposes an input 3D point cloud into a small set of learned prototypical 3D shapes. The resulting reconstruction is visually interpretable and can be used to perform unsupervised instance and low-shot semantic segmentation of complex scenes. We demonstrate the usefulness of our model on a novel dataset of seven large aerial LiDAR scans from diverse real-world scenarios. Our approach outperforms state-of-the-art unsupervised methods in terms of decomposition accuracy while remaining visually interpretable. Our code and dataset are available at https://romainloiseau.fr/learnable-earth-parser/
- Published
- 2023
39. Pixel-wise Agricultural Image Time Series Classification: Comparisons and a Deformable Prototype-based Approach
- Author
-
Vincent, Elliot, Ponce, Jean, and Aubry, Mathieu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Improvements in Earth observation by satellites allow for imagery of ever higher temporal and spatial resolution. Leveraging this data for agricultural monitoring is key for addressing environmental and economic challenges. Current methods for crop segmentation using temporal data either rely on annotated data or are heavily engineered to compensate the lack of supervision. In this paper, we present and compare datasets and methods for both supervised and unsupervised pixel-wise segmentation of satellite image time series (SITS). We also introduce an approach to add invariance to spectral deformations and temporal shifts to classical prototype-based methods such as K-means and Nearest Centroid Classifier (NCC). We study different levels of supervision and show this simple and highly interpretable method achieves the best performance in the low data regime and significantly improves the state of the art for unsupervised classification of agricultural time series on four recent SITS datasets., Comment: Revised version. Added references and baselines. Corrected typos. Added discussion section and Appendix A, B and C
- Published
- 2023
40. Three-Dimensional Ultrasound Matrix Imaging
- Author
-
Bureau, Flavien, Robin, Justine, Ber, Arthur Le, Lambert, William, Fink, Mathias, and Aubry, Alexandre
- Subjects
Electrical Engineering and Systems Science - Image and Video Processing ,Physics - Applied Physics ,Physics - Medical Physics - Abstract
Matrix imaging paves the way towards a next revolution in wave physics. Based on the response matrix recorded between a set of sensors, it enables an optimized compensation of aberration phenomena and multiple scattering events that usually drastically hinder the focusing process in heterogeneous media. Although it gave rise to spectacular results in optical microscopy or seismic imaging, the success of matrix imaging has been so far relatively limited with ultrasonic waves because wave control is generally only performed with a linear array of transducers. In this paper, we extend ultrasound matrix imaging to a 3D geometry. Switching from a 1D to a 2D probe enables a much sharper estimation of the transmission matrix that links each transducer and each medium voxel. Here, we first present an experimental proof of concept on a tissue-mimicking phantom through ex-vivo tissues and then, show the potential of 3D matrix imaging for transcranial applications., Comment: 60 pages, 14 figures
- Published
- 2023
- Full Text
- View/download PDF
41. Imaging the crustal and upper mantle structure of the North Anatolian Fault: A Transmission Matrix Framework for Local Adaptive Focusing
- Author
-
Touma, Rita, Ber, Arthur Le, Campillo, Michel, and Aubry, Alexandre
- Subjects
Physics - Geophysics - Abstract
Imaging the structure of major fault zones is essential for our understanding of crustal deformations and their implications on seismic hazards. Investigating such complex regions presents several issues, including the variation of seismic velocity due to the diversity of geological units and the cumulative damage caused by earthquakes. Conventional migration techniques are in general strongly sensitive to the available velocity model. Here we apply a passive matrix imaging approach which is robust to the mismatch between this model and the real seismic velocity distribution. This method relies on the cross-correlation of ambient noise recorded by a geophone array. The resulting set of impulse responses form a reflection matrix that contains all the information about the subsurface. In particular, the reflected body waves can be leveraged to: (i) determine the transmission matrix between the Earth's surface and any point in the subsurface; (ii) build a confocal image of the subsurface reflectivity with a transverse resolution only limited by diffraction. As a study case, we consider seismic noise (0.1-0.5 Hz) recorded by the Dense Array for Northern Anatolia (DANA) that consists of 73 stations deployed for 18 months in the region of the 1999 Izmit earthquake. Passive matrix imaging reveals the scattering structure of the crust and upper mantle around the NAFZ over a depth range of 60 km. The results show that most of the scattering is associated with the Northern branch that passes throughout the crust and penetrates into the upper mantle., Comment: 44 pages, 9 figures
- Published
- 2023
42. Harnessing Forward Multiple Scattering for Optical Imaging Deep Inside an Opaque Medium
- Author
-
Najar, Ulysse, Barolle, Victor, Balondrade, Paul, Fink, Mathias, Boccara, A. Claude, and Aubry, Alexandre
- Subjects
Physics - Optics - Abstract
As light travels through a disordered medium such as biological tissues, it undergoes multiple scattering events. This phenomenon is detrimental to in-depth optical microscopy, as it causes a drastic degradation of contrast, resolution and brightness of the resulting image beyond a few scattering mean free paths. However, the information about the inner reflectivity of the sample is not lost; only scrambled. To recover this information, a matrix approach of optical imaging can be fruitful. Here, we report on a de-scanned measurement of a high-dimension reflection matrix R via low coherence interferometry. Then, we show how a set of independent focusing laws can be extracted for each medium voxel through an iterative multi-scale analysis of wave distortions contained in R. It enables an optimal and local compensation of forward multiple scattering paths and provides a three-dimensional confocal image of the sample as the latter one had become digitally transparent. The proof-of-concept experiment is performed on a human opaque cornea and an extension of the penetration depth by a factor five is demonstrated compared to the state-of-the-art., Comment: 78 pages, 29 figures, 1 table
- Published
- 2023
- Full Text
- View/download PDF
43. Bandgap fluctuations and robustness in two-dimensional hyperuniform dielectric materials
- Author
-
Froufe-Pérez, Luis S., Aubry, Geoffroy, Scheffold, Frank, and Magkiriadou, Sofia
- Subjects
Physics - Classical Physics ,Physics - Optics - Abstract
We numerically study the statistical fluctuations of photonic band gaps in ensembles of stealthy hyperuniform disordered patterns. We find that at low stealthiness, where correlations are weak, band gaps of different system realizations appear over a wide frequency range, are narrow, and generally do not overlap. Interestingly, above a critical value of stealthiness $\chi \gtrsim 0.35$, the bandgaps become large and overlap significantly from realization to realization, while a second gap appears. These observations extend our understanding of photonic bandgaps in disordered systems and provide information on the robustness of gaps in practical applications., Comment: 7 pages, 4 figures. Luis S. Froufe-P\'erez, Geoffroy Aubry, and Frank Scheffold have equal first authorship
- Published
- 2023
- Full Text
- View/download PDF
44. The Learnable Typewriter: A Generative Approach to Text Analysis
- Author
-
Siglidis, Ioannis, Gonthier, Nicolas, Gaubil, Julien, Monnier, Tom, and Aubry, Mathieu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
We present a generative document-specific approach to character analysis and recognition in text lines. Our main idea is to build on unsupervised multi-object segmentation methods and in particular those that reconstruct images based on a limited amount of visual elements, called sprites. Taking as input a set of text lines with similar font or handwriting, our approach can learn a large number of different characters and leverage line-level annotations when available. Our contribution is twofold. First, we provide the first adaptation and evaluation of a deep unsupervised multi-object segmentation approach for text line analysis. Since these methods have mainly been evaluated on synthetic data in a completely unsupervised setting, demonstrating that they can be adapted and quantitatively evaluated on real images of text and that they can be trained using weak supervision are significant progresses. Second, we show the potential of our method for new applications, more specifically in the field of paleography, which studies the history and variations of handwriting, and for cipher analysis. We demonstrate our approach on three very different datasets: a printed volume of the Google1000 dataset, the Copiale cipher and historical handwritten charters from the 12th and early 13th century., Comment: For the code and a quick-overview visit the project webpage at http://imagine.enpc.fr/~siglidii/learnable-typewriter
- Published
- 2023
45. Are inertial vacua equivalent in Lorentz-violating theories? Does it matter?
- Author
-
Costa, Bruno Arderucio, Bonder, Yuri, and Juárez-Aubry, Benito A.
- Subjects
High Energy Physics - Theory ,General Relativity and Quantum Cosmology - Abstract
Several approaches to quantum gravity suggest violations of Lorentz symmetry as low-energy signatures. This article uses a concrete Lorentz-violating quantum field theory to study different inertial vacua. We show that they are unitarily inequivalent and that the vacuum in one inertial frame appears, in a different inertial frame, to be populated with particles of arbitrarily high momenta. At first sight, this poses a critical challenge to the physical validity of Lorentz-violating theories, since we do not witness vacuum excitations by changing inertial frames. Nevertheless, we demonstrate that inertial Unruh-De Witt detectors are insensitive to these effects. We also discuss the Hadamard condition for this Lorentz-violating theory., Comment: Revised and extended version. 12 pages, 1 figure
- Published
- 2023
- Full Text
- View/download PDF
46. Large Deviations for Classification Performance Analysis of Machine Learning Systems
- Author
-
Braca, Paolo, Millefiori, Leonardo M., Aubry, Augusto, De Maio, Antonio, and Willett, Peter
- Subjects
Computer Science - Machine Learning ,Computer Science - Artificial Intelligence ,Electrical Engineering and Systems Science - Signal Processing ,Mathematics - Probability ,Statistics - Applications ,Statistics - Machine Learning - Abstract
We study the performance of machine learning binary classification techniques in terms of error probabilities. The statistical test is based on the Data-Driven Decision Function (D3F), learned in the training phase, i.e., what is thresholded before the final binary decision is made. Based on large deviations theory, we show that under appropriate conditions the classification error probabilities vanish exponentially, as $\sim \exp\left(-n\,I + o(n) \right)$, where $I$ is the error rate and $n$ is the number of observations available for testing. We also propose two different approximations for the error probability curves, one based on a refined asymptotic formula (often referred to as exact asymptotics), and another one based on the central limit theorem. The theoretical findings are finally tested using the popular MNIST dataset., Comment: 5 pages, 3 figures, 1 table
- Published
- 2023
47. MegaPose: 6D Pose Estimation of Novel Objects via Render & Compare
- Author
-
Labbé, Yann, Manuelli, Lucas, Mousavian, Arsalan, Tyree, Stephen, Birchfield, Stan, Tremblay, Jonathan, Carpentier, Justin, Aubry, Mathieu, Fox, Dieter, and Sivic, Josef
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Robotics - Abstract
We introduce MegaPose, a method to estimate the 6D pose of novel objects, that is, objects unseen during training. At inference time, the method only assumes knowledge of (i) a region of interest displaying the object in the image and (ii) a CAD model of the observed object. The contributions of this work are threefold. First, we present a 6D pose refiner based on a render&compare strategy which can be applied to novel objects. The shape and coordinate system of the novel object are provided as inputs to the network by rendering multiple synthetic views of the object's CAD model. Second, we introduce a novel approach for coarse pose estimation which leverages a network trained to classify whether the pose error between a synthetic rendering and an observed image of the same object can be corrected by the refiner. Third, we introduce a large-scale synthetic dataset of photorealistic images of thousands of objects with diverse visual and shape properties and show that this diversity is crucial to obtain good generalization performance on novel objects. We train our approach on this large synthetic dataset and apply it without retraining to hundreds of novel objects in real images from several pose estimation benchmarks. Our approach achieves state-of-the-art performance on the ModelNet and YCB-Video datasets. An extensive evaluation on the 7 core datasets of the BOP challenge demonstrates that our approach achieves performance competitive with existing approaches that require access to the target objects during training. Code, dataset and trained models are available on the project page: https://megapose6d.github.io/., Comment: CoRL 2022
- Published
- 2022
48. Diffusion without Spreading of a Wave Packet in Nonlinear Random Models
- Author
-
Aubry, Serge J.
- Subjects
Condensed Matter - Disordered Systems and Neural Networks ,Mathematical Physics ,Nonlinear Sciences - Chaotic Dynamics - Abstract
We discuss the long time behaviour of a finite energy wave packet in nonlinear Hamiltonians on infinite lattices at arbitrary dimension, exhibiting linear Anderson localization. Strong arguments both mathematical and numerical, suggest for infinite models that small amplitude wave packets may generate stationary quasiperiodic solutions (KAM tori) almost undistinguishable from linear wave packets. The probability of this event is non vanishing at small enough amplitude and goes to unity at amplitude zero. Most other wave packets (non KAM tori) are chaotic. We discuss the Arnold diffusion conjecture (recently partially proven) and propose a modified Boltzmann statistics for wave packets valid in generic models. The consequence is that the probability that a chaotic wave packet spreads to zero amplitude is zero. It must always remain focused around one or few chaotic spots which moves randomly over the whole system and generates subdiffusion.We study a class of Ding Dong models also generating subdiffusion where the nonlinearities are replaced by hard core potentials. Then we prove rigorously that spreading is impossible for any initial wave packet., Comment: 27 pages, 1 figure, Proceedings of the 28th School-Conference, Chania, 2022
- Published
- 2022
49. Adaptive Radar Detection and Bearing Estimation in the Presence of Unknown Mutual Coupling
- Author
-
Aubry, Augusto, De Maio, Antonio, Lan, Lan, and Rosamilia, Massimo
- Subjects
Electrical Engineering and Systems Science - Signal Processing - Abstract
This paper deals with joint adaptive radar detection and target bearing estimation in the presence of mutual coupling among the array elements. First of all, a suitable model of the signal received by the multichannel radar is developed via a linearization procedure of the Uniform Linear Array (ULA) manifold around the nominal array looking direction together with the use of symmetric Toeplitz structured matrices to represent the mutual coupling effects. Hence, the Generalized Likelihood Ratio Test (GLRT) detector is evaluated under the assumption of homogeneous radar environment. Its computation leverages a specific Minorization-Maximization (MM) framework, with proven convergence properties, to optimize the concentrated likelihood function under the target presence hypothesis. Besides, when the number of active mutual coupling coefficients is unknown, a Multifamily Likelihood Ratio Test (MFLRT) approach is invoked. During the analysis phase, the performance of the new detectors is compared with benchmarks as well as with counterparts available in the open literature which neglect the mutual coupling phenomenon. The results indicate that it is necessary to consider judiciously the coupling effect since the design phase, to guarantee performance levels close to the benchmark., Comment: submitted to IEEE Transactions on Signal Processing
- Published
- 2022
- Full Text
- View/download PDF
50. Enhanced mobility of dislocation network nodes and its effect on dislocation multiplication and strain hardening
- Author
-
Bertin, Nicolas, Cai, Wei, Aubry, Sylvie, Arsenlis, Athanasios, and Bulatov, Vasily V.
- Subjects
Condensed Matter - Materials Science - Abstract
Understanding plastic deformation of crystals in terms of the fundamental physics of dislocations has remained a grand challenge in materials science for decades. To overcome this, the Discrete Dislocation Dynamics (DDD) method has been developed, but its lack of atomistic resolution leaves open the possibility that certain key mechanisms may be overlooked. By comparing large-scale Molecular Dynamics (MD) with DDD simulations performed under identical conditions we uncover significant discrepancies in the predicted strength and microstructure evolution in BCC crytals under high-strain rate conditions. These are traced to unexpected behaviors of dislocation network nodes forming at dislocation intersections, that can move in ways not previously anticipated as revealed by MD. Once these newfound freedoms of nodal motion are incorporated, DDD simulations begin to closely match plastic evolution observed in MD. This additional mechanism of motion whereby non-screw dislocations can change their glide plane profoundly affects fundamental processes of dislocation multiplication, recovery and storage that define strength of metals.
- Published
- 2022
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.