Search

Underwater noise pollution from human activities, particularly shipping, has been recognised as a serious threat to marine life. The sound generated by vessels can have various adverse effects on fish and aquatic ecosystems in general. In this setting, the estimation and analysis of the underwater noise produced by vessels is an important challenge for the preservation of the marine environment. In this paper we propose a model for the spatio-temporal characterisation of the underwater noise generated by vessels. The approach is based on the reconstruction of the vessels' trajectories from Automatic Identification System (AIS) data and on their deployment in a spatio-temporal database. Trajectories are enriched with semantic information like the acoustic characteristics of the vessels' engines or the activity performed by the vessels. We define a model for underwater noise propagation and use the trajectories' information to infer how noise propagates in the area of interest. We develop our approach for the case study of the fishery activities in the Northern Adriatic sea, an area of the Mediterranean sea which is well known to be highly exploited. We implement our approach using MobilityDB, an open source geospatial trajectory data management and analysis platform, which offers spatio-temporal operators and indexes improving the efficiency of our system. We use this platform to conduct various analyses of the underwater noise generated in the Northern Adriatic Sea, aiming at estimating the impact of fishing activities on underwater noise pollution and at demonstrating the flexibility and expressiveness of our approach.

MPS-VG, a Virtual Grating (VG) model for the Lagrangian mesh-free Moving Particle Semi-implicit (MPS) method is proposed for replacing conventional particle-based solid modeling of gratings with a set of thin inclined slats. Unlike most approaches for perforated wave energy dampening devices, in which the flow through the device is simplified by pressure loss or damping effects without flow deflection, MPS-VG models the angular deviation caused by hydrodynamic impact on inclined slats. Both accuracy and computational performance of the model were checked through a simulation of wet dam break scenarios with the grating structures placed horizontally or vertically. The results were compared with those from fully particle-based modeling. MPS-VG correctly predicted complex wave-structure interactions using a relatively low-resolution model and significantly reduced processing time and memory storage compared to conventional particle-based MPS modeling. The evaluation of the performance of the gratings with inclined slats as wave energy dampers revealed the horizontal gratings outperformed the vertical ones. Therefore, qualitative and quantitative agreements strengthened the potential of MPS-VG as a practical and computationally efficient tool for the study of multi-scale phenomena of wave impacts on grating with inclined slats., Comment: 39, 26 figures

Self-supervised visual foundation models produce powerful embeddings that achieve remarkable performance on a wide range of downstream tasks. However, unlike vision-language models such as CLIP, self-supervised visual features are not readily aligned with language, hindering their adoption in open-vocabulary tasks. Our method, named dino.txt, unlocks this new ability for DINOv2, a widely used self-supervised visual encoder. We build upon the LiT training strategy, which trains a text encoder to align with a frozen vision model but leads to unsatisfactory results on dense tasks. We propose several key ingredients to improve performance on both global and dense tasks, such as concatenating the [CLS] token with the patch average to train the alignment and curating data using both text and image modalities. With these, we successfully train a CLIP-like model with only a fraction of the computational cost compared to CLIP while achieving state-of-the-art results in zero-shot classification and open-vocabulary semantic segmentation.

This letter reports on Solidly-Mounted Bidimensional Mode Resonators (S2MRs) based on 30% Scandium-doped Aluminum Nitride (ScAlN) on Silicon Carbide (SiC), operating near 16 GHz. Experimental results demonstrate mechanical quality factors (Qm) as high as 380, electromechanical coupling coefficients (kt2) of 4.5%, an overall Figure of Merit (FOM = Qmkt2) exceeding 17, and power handling greater than 20 dBm for devices closely matched to 50 ohm. To the best of the authors' knowledge, S2MRs exhibit the highest Key Performance Indicators (KPIs) among solidly mounted resonators in the Ku band, paving the way for the integration of nanoacoustic devices on fast substrates with high-power electronics, tailored for military and harsh environment applications., Comment: Submitted to IEEE EDL

Vision Language Models (VLMs) have demonstrated remarkable capabilities in various open-vocabulary tasks, yet their zero-shot performance lags behind task-specific finetuned models, particularly in complex tasks like Referring Expression Comprehension (REC). Fine-tuning usually requires 'white-box' access to the model's architecture and weights, which is not always feasible due to proprietary or privacy concerns. In this work, we propose LLM-wrapper, a method for 'black-box' adaptation of VLMs for the REC task using Large Language Models (LLMs). LLM-wrapper capitalizes on the reasoning abilities of LLMs, improved with a light fine-tuning, to select the most relevant bounding box matching the referring expression, from candidates generated by a zero-shot black-box VLM. Our approach offers several advantages: it enables the adaptation of closed-source models without needing access to their internal workings, it is versatile as it works with any VLM, it transfers to new VLMs, and it allows for the adaptation of an ensemble of VLMs. We evaluate LLM-wrapper on multiple datasets using different VLMs and LLMs, demonstrating significant performance improvements and highlighting the versatility of our method. While LLM-wrapper is not meant to directly compete with standard white-box fine-tuning, it offers a practical and effective alternative for black-box VLM adaptation. The code will be open-sourced., Comment: v1 at EVAL-FoMo workshop, ECCV 2024

Annotating lidar point clouds for autonomous driving is a notoriously expensive and time-consuming task. In this work, we show that the quality of recent self-supervised lidar scan representations allows a great reduction of the annotation cost. Our method has two main steps. First, we show that self-supervised representations allow a simple and direct selection of highly informative lidar scans to annotate: training a network on these selected scans leads to much better results than a random selection of scans and, more interestingly, to results on par with selections made by SOTA active learning methods. In a second step, we leverage the same self-supervised representations to cluster points in our selected scans. Asking the annotator to classify each cluster, with a single click per cluster, then permits us to close the gap with fully-annotated training sets, while only requiring one thousandth of the point labels.

Recent CLIP-like Vision-Language Models (VLMs), pre-trained on large amounts of image-text pairs to align both modalities with a simple contrastive objective, have paved the way to open-vocabulary semantic segmentation. Given an arbitrary set of textual queries, image pixels are assigned the closest query in feature space. However, this works well when a user exhaustively lists all possible visual concepts in an image, which contrast against each other for the assignment. This corresponds to the current evaluation setup in the literature which relies on having access to a list of in-domain relevant concepts, typically classes of a benchmark dataset. Here, we consider the more challenging (and realistic) scenario of segmenting a single concept, given a textual prompt and nothing else. To achieve good results, besides contrasting with the generic $\textit{background}$ text, we propose two different approaches to automatically generate, at test time, textual contrastive concepts that are query-specific. We do so by leveraging the distribution of text in the VLM's training set or crafted LLM prompts. We also propose a metric designed to evaluate this scenario and show the relevance of our approach to commonly used datasets.

This work reports an acoustic solidly mounted resonator (SMR) at 18.64 GHz, among the highest operating frequencies reported. The device is built in scandium aluminum nitride (ScAlN) on top of silicon dioxide (SiO2) and tantalum pentoxide (Ta2O5) Bragg reflectors on silicon (Si) wafer. The stack is analyzed with X-ray reflectivity (XRR) and high-resolution X-ray diffraction (HRXRD). The resonator shows a coupling coefficient (k2) of 2.0%, high series quality factor (Qs) of 156, shunt quality factor (Qp) of 142, and maximum Bode quality factor (Qmax) of 210. The third-order harmonics at 59.64 GHz is also observed with k2 around 0.6% and Q around 40. Upon further development, the reported acoustic resonator platform can enable various front-end signal-processing functions, e.g., filters and oscillators, at future frequency range 3 (FR3) bands., Comment: 5 pages, 9 figures, 5 tables

Motion forecasting is crucial in autonomous driving systems to anticipate the future trajectories of surrounding agents such as pedestrians, vehicles, and traffic signals. In end-to-end forecasting, the model must jointly detect and track from sensor data (cameras or LiDARs) the past trajectories of the different elements of the scene and predict their future locations. We depart from the current trend of tackling this task via end-to-end training from perception to forecasting, and instead use a modular approach. We individually build and train detection, tracking and forecasting modules. We then only use consecutive finetuning steps to integrate the modules better and alleviate compounding errors. We conduct an in-depth study on the finetuning strategies and it reveals that our simple yet effective approach significantly improves performance on the end-to-end forecasting benchmark. Consequently, our solution ranks first in the Argoverse 2 End-to-end Forecasting Challenge, with 63.82 mAPf. We surpass forecasting results by +17.1 points over last year's winner and by +13.3 points over this year's runner-up. This remarkable performance in forecasting can be explained by our modular paradigm, which integrates finetuning strategies and significantly outperforms the end-to-end-trained counterparts. The code, model weights and results are made available https://github.com/valeoai/valeo4cast., Comment: Winning solution of the Argoverse 2 "Unified Detection, Tracking, and Forecasting" challenge; work accepted at Road++ ECCVW 2024

Aims: We address two issues for the adoption of convex optimization in radio interferometric imaging. First, a method for a fine resolution setup is proposed which scales naturally in terms of memory usage and reconstruction speed. Second, a new tool to localize a region of uncertainty is developed, paving the way for quantitative imaging in radio interferometry. Methods: The classical $\ell_1$ penalty is used to turn the inverse problem into a sparsity-promoting optimization. For efficient implementation, the so-called Frank-Wolfe algorithm is used together with a \textit{polyatomic} refinement. The algorithm naturally produces sparse images at each iteration, leveraged to reduce memory and computational requirements. In that regard, PolyCLEAN reproduces the numerical behavior of CLEAN while guaranteeing that it solves the minimization problem of interest. Additionally, we introduce the concept of the \textit{dual certificate image}, which appears as a numerical byproduct of the Frank-Wolfe algorithm. This image is proposed as a tool for uncertainty quantification on the location of the recovered sources. Results: PolyCLEAN demonstrates good scalability performance, in particular for fine-resolution grids. On simulations, the Python-based implementation is competitive with the fast numerically-optimized CLEAN solver. This acceleration does not affect image reconstruction quality: PolyCLEAN images are consistent with CLEAN-obtained ones for both point sources and diffuse emission recovery. We also highlight PolyCLEAN reconstruction capabilities on observed radio measurements. Conclusions: PolyCLEAN can be considered as an alternative to CLEAN in the radio interferometric imaging pipeline, as it enables the use of Bayesian priors without impacting the scalability and numerical performance of the imaging method.

We investigate the dynamics of plasma-based acceleration processes with collisionless particle dynamics and non negligible thermal effects. We aim at assessing the applicability of fluid-like models, obtained by suitable closure assumptions applied to the relativistic kinetic equations, thus not suffering of statistical noise, even in presence of a finite temperature. The work here presented focuses on the characterization of pressure anisotropies, which crucially depend on the adopted closure scheme, and hence are useful to discern the appropriate thermal fluid model. To this aim, simulation results of spatially resolved fluid models with different thermal closure assumptions are compared with the results of particle-in-cell (PIC) simulations at changing temperature and amplitude of plasma oscillations., Comment: The following article has been submitted to Physics of Plasma

This work reports on the modeling, fabrication, and experimental characterization of a 13 GHz 30% Scandium-doped Aluminum Nitride (ScAlN) Overmoded Bulk Acoustic Resonator (OBAR) for high-frequency Radio Frequency (RF) applications, notably in 5G technology and beyond. The Finite Element Analysis (FEA) optimization process targets the top and bottom metal electrode thicknesses, balancing the electromechanical coupling coefficient and acoustic energy distribution to enhance device Figure of Merit (FOM). Experimental results on fabricated devices employing platinum and aluminum as bottom and top electrode, respectively, demonstrate a quality factor at resonance (Qs) of 210 and a coupling coefficient (kt2) of 5.2% at 13.3 GHz for the second bulk thickness overtone, effectively validating the simulation framework and hinting at the possible implementation of OBARs for advanced RF filters in 5G networks., Comment: 4 pages, 4 figures, submitted to 2024 Hilton Head Workshop Proceedings

The popular CLIP model displays impressive zero-shot capabilities thanks to its seamless interaction with arbitrary text prompts. However, its lack of spatial awareness makes it unsuitable for dense computer vision tasks, e.g., semantic segmentation, without an additional fine-tuning step that often uses annotations and can potentially suppress its original open-vocabulary properties. Meanwhile, self-supervised representation methods have demonstrated good localization properties without human-made annotations nor explicit supervision. In this work, we take the best of both worlds and propose an open-vocabulary semantic segmentation method, which does not require any annotations. We propose to locally improve dense MaskCLIP features, which are computed with a simple modification of CLIP's last pooling layer, by integrating localization priors extracted from self-supervised features. By doing so, we greatly improve the performance of MaskCLIP and produce smooth outputs. Moreover, we show that the used self-supervised feature properties can directly be learnt from CLIP features. Our method CLIP-DINOiser needs only a single forward pass of CLIP and two light convolutional layers at inference, no extra supervision nor extra memory and reaches state-of-the-art results on challenging and fine-grained benchmarks such as COCO, Pascal Context, Cityscapes and ADE20k. The code to reproduce our results is available at https://github.com/wysoczanska/clip_dinoiser.

This work describes sputtered scandium aluminum nitride (ScAlN) thin-film bulk acoustic resonators (FBAR) at millimeter wave (mmWave) with high quality factor (Q) using platinum (Pt) electrodes. FBARs with combinations of Pt and aluminum (Al) electrodes, i.e., Al top Al bottom, Pt top Al bottom, Al top Pt bottom, and Pt top Pt bottom, are built to study the impact of electrodes on mmWave FBARs. The demonstrated FBAR with Pt top and bottom electrodes achieve electromechanical coupling (k2) of 4.0% and Q of 116 for the first-order symmetric (S1) mode at 13.7 GHz, and k2 of 1.8% and Q of 94 for third-order symmetric (S3) mode at 61.6 GHz. Through these results, we confirmed that even in the frequency band of approximately 60 GHz, ScAlN FBAR can achieve a Q factor approaching 100 with optimized fabrication and acoustic/EM design. Further development calls for stacks with better quality in piezoelectric and metallic layers., Comment: 4 pages, 7 figures, accepted by IEEE MEMS 2024

Self-supervised image backbones can be used to address complex 2D tasks (e.g., semantic segmentation, object discovery) very efficiently and with little or no downstream supervision. Ideally, 3D backbones for lidar should be able to inherit these properties after distillation of these powerful 2D features. The most recent methods for image-to-lidar distillation on autonomous driving data show promising results, obtained thanks to distillation methods that keep improving. Yet, we still notice a large performance gap when measuring the quality of distilled and fully supervised features by linear probing. In this work, instead of focusing only on the distillation method, we study the effect of three pillars for distillation: the 3D backbone, the pretrained 2D backbones, and the pretraining dataset. In particular, thanks to our scalable distillation method named ScaLR, we show that scaling the 2D and 3D backbones and pretraining on diverse datasets leads to a substantial improvement of the feature quality. This allows us to significantly reduce the gap between the quality of distilled and fully-supervised 3D features, and to improve the robustness of the pretrained backbones to domain gaps and perturbations., Comment: The code is available at https://github.com/valeoai/ScaLR

The recent enthusiasm for open-world vision systems show the high interest of the community to perform perception tasks outside of the closed-vocabulary benchmark setups which have been so popular until now. Being able to discover objects in images/videos without knowing in advance what objects populate the dataset is an exciting prospect. But how to find objects without knowing anything about them? Recent works show that it is possible to perform class-agnostic unsupervised object localization by exploiting self-supervised pre-trained features. We propose here a survey of unsupervised object localization methods that discover objects in images without requiring any manual annotation in the era of self-supervised ViTs. We gather links of discussed methods in the repository https://github.com/valeoai/Awesome-Unsupervised-Object-Localization., Comment: IJCV 2024

The Bluebild algorithm is a new technique for image synthesis in radio astronomy which decomposes the sky into distinct energy levels using functional principal component analysis. These levels can be linearly combined to construct a least-squares estimate of the radio sky, i.e. minimizing the residuals between measured and predicted visibilities. This approach is particularly useful for deconvolution-free imaging or for scientific applications that need to filter specific energy levels. We present an HPC implementation of the Bluebild algorithm for radio-interferometric imaging: Bluebild Imaging++ (BIPP). The library features interfaces to C++, C and Python and is designed with seamless GPU acceleration in mind. We evaluate the accuracy and performance of BIPP on simulated observations of the upcoming Square Kilometer Array Observatory and real data from the Low-Frequency Array (LOFAR) telescope. We find that BIPP offers accurate wide-field imaging and has competitive execution time with respect to the interferometric imaging libraries CASA and WSClean for images with $\leq 10^6$ pixels. Furthermore, due to the energy level decomposition, images produced with BIPP can reveal information about faint and diffuse structures before any cleaning iterations. BIPP does not perform any regularization, but we suggest methods to integrate the output of BIPP with CLEAN. The source code of BIPP is publicly released., Comment: 21 pages, 13 figures

The emergence of CLIP has opened the way for open-world image perception. The zero-shot classification capabilities of the model are impressive but are harder to use for dense tasks such as image segmentation. Several methods have proposed different modifications and learning schemes to produce dense output. Instead, we propose in this work an open-vocabulary semantic segmentation method, dubbed CLIP-DIY, which does not require any additional training or annotations, but instead leverages existing unsupervised object localization approaches. In particular, CLIP-DIY is a multi-scale approach that directly exploits CLIP classification abilities on patches of different sizes and aggregates the decision in a single map. We further guide the segmentation using foreground/background scores obtained using unsupervised object localization methods. With our method, we obtain state-of-the-art zero-shot semantic segmentation results on PASCAL VOC and perform on par with the best methods on COCO.

A comprehensive characterization of lattice Boltzmann (LB) schemes to perform warm fluid numerical simulations of particle wakefield acceleration (PWFA) processes is discussed in this paper. The LB schemes we develop hinge on the moment matching procedure, allowing the fluid description of a warm relativistic plasma wake generated by a driver pulse propagating in a neutral plasma. We focus on fluid models equations resulting from two popular closure assumptions of the relativistic kinetic equations, i.e., the local equilibrium and the warm plasma closure assumptions. The developed LB schemes can thus be used to disclose insights on the quantitative differences between the two closure approaches in the dynamics of PWFA processes. Comparisons between the proposed schemes and available analytical results are extensively addressed., Comment: Published on Physics of Plasmas

This work reports a millimeter wave (mmWave) thin-film bulk acoustic resonator (FBAR) in sputtered scandium aluminum nitride (ScAlN). This paper identifies challenges of frequency scaling sputtered ScAlN into mmWave and proposes a stack and new fabrication procedure with a sputtered Sc0.3Al0.7N on Al on Si carrier wafer. The resonator achieves electromechanical coupling (k2) of 7.0% and quality factor (Q) of 62 for the first-order symmetric (S1) mode at 21.4 GHz, along with k2 of 4.0% and Q of 19 for the third-order symmetric (S3) mode at 55.4 GHz, showing higher figures of merit (FoM, k2xQ) than reported AlN/ScAlN-based mmWave acoustic resonators. The ScAlN quality is identified by transmission electron microscopy (TEM) and X-ray diffraction (XRD), identifying the bottlenecks in the existing piezoelectric-metal stack. Further improvement of ScAlN/AlN-based mmWave acoustic resonators calls for better crystalline quality from improved thin-film deposition methods., Comment: 3 pages, 7 figures, submitted to JMEMS

Self-supervised learning can be used for mitigating the greedy needs of Vision Transformer networks for very large fully-annotated datasets. Different classes of self-supervised learning offer representations with either good contextual reasoning properties, e.g., using masked image modeling strategies, or invariance to image perturbations, e.g., with contrastive methods. In this work, we propose a single-stage and standalone method, MOCA, which unifies both desired properties using novel mask-and-predict objectives defined with high-level features (instead of pixel-level details). Moreover, we show how to effectively employ both learning paradigms in a synergistic and computation-efficient way. Doing so, we achieve new state-of-the-art results on low-shot settings and strong experimental results in various evaluation protocols with a training that is at least 3 times faster than prior methods. We provide the implementation code at https://github.com/valeoai/MOCA., Comment: Published in Transactions on Machine Learning Research (TMLR) 2024. Code at https://github.com/valeoai/MOCA

Analysis and synthesis are key steps of the radio-interferometric imaging process, serving as a bridge between visibility and sky domains. They can be expressed as partial Fourier transforms involving a large number of non-uniform frequencies and spherically-constrained spatial coordinates. Due to the data non-uniformity, these partial Fourier transforms are computationally expensive and represent a serious bottleneck in the image reconstruction process. The W-gridding algorithm achieves log-linear complexity for both steps by applying a series of 2D non-uniform FFTs (NUFFT) to the data sliced along the so-called $w$ frequency coordinate. A major drawback of this method however is its restriction to direction-cosine meshes, which are fundamentally ill-suited for large field of views. This paper introduces the HVOX gridder, a novel algorithm for analysis/synthesis based on a 3D-NUFFT. Unlike W-gridding, the latter is compatible with arbitrary spherical meshes such as the popular HEALPix scheme for spherical data processing. The 3D-NUFFT allows one to optimally select the size of the inner FFTs, in particular the number of W-planes. This results in a better performing and auto-tuned algorithm, with controlled accuracy guarantees backed by strong results from approximation theory. To cope with the challenging scale of next-generation radio telescopes, we propose moreover a chunked evaluation strategy: by partitioning the visibility and sky domains, the 3D-NUFFT is decomposed into sub-problems which execute in parallel, while simultaneously cutting memory requirements. Our benchmarking results demonstrate the scalability of HVOX for both SKA and LOFAR, considering state-of-the-art challenging imaging setups. HVOX is moreover computationally competitive with W-gridder, despite the absence of domain-specific optimizations in our implementation.

We propose SeedAL, a method to seed active learning for efficient annotation of 3D point clouds for semantic segmentation. Active Learning (AL) iteratively selects relevant data fractions to annotate within a given budget, but requires a first fraction of the dataset (a 'seed') to be already annotated to estimate the benefit of annotating other data fractions. We first show that the choice of the seed can significantly affect the performance of many AL methods. We then propose a method for automatically constructing a seed that will ensure good performance for AL. Assuming that images of the point clouds are available, which is common, our method relies on powerful unsupervised image features to measure the diversity of the point clouds. It selects the point clouds for the seed by optimizing the diversity under an annotation budget, which can be done by solving a linear optimization problem. Our experiments demonstrate the effectiveness of our approach compared to random seeding and existing methods on both the S3DIS and SemanticKitti datasets. Code is available at https://github.com/nerminsamet/seedal., Comment: ICCV 2023

Recent advances in self-supervised visual representation learning have paved the way for unsupervised methods tackling tasks such as object discovery and instance segmentation. However, discovering objects in an image with no supervision is a very hard task; what are the desired objects, when to separate them into parts, how many are there, and of what classes? The answers to these questions depend on the tasks and datasets of evaluation. In this work, we take a different approach and propose to look for the background instead. This way, the salient objects emerge as a by-product without any strong assumption on what an object should be. We propose FOUND, a simple model made of a single $conv1\times1$ initialized with coarse background masks extracted from self-supervised patch-based representations. After fast training and refining these seed masks, the model reaches state-of-the-art results on unsupervised saliency detection and object discovery benchmarks. Moreover, we show that our approach yields good results in the unsupervised semantic segmentation retrieval task. The code to reproduce our results is available at https://github.com/valeoai/FOUND., Comment: CVPR 2023

Lensless imaging can provide visual privacy due to the highly multiplexed characteristic of its measurements. However, this alone is a weak form of security, as various adversarial attacks can be designed to invert the one-to-many scene mapping of such cameras. In this work, we enhance the privacy provided by lensless imaging by (1) downsampling at the sensor and (2) using a programmable mask with variable patterns as our optical encoder. We build a prototype from a low-cost LCD and Raspberry Pi components, for a total cost of around 100 USD. This very low price point allows our system to be deployed and leveraged in a broad range of applications. In our experiments, we first demonstrate the viability and reconfigurability of our system by applying it to various classification tasks: MNIST, CelebA (face attributes), and CIFAR10. By jointly optimizing the mask pattern and a digital classifier in an end-to-end fashion, low-dimensional, privacy-enhancing embeddings are learned directly at the sensor. Secondly, we show how the proposed system, through variable mask patterns, can thwart adversaries that attempt to invert the system (1) via plaintext attacks or (2) in the event of camera parameters leaks. We demonstrate the defense of our system to both risks, with 55% and 26% drops in image quality metrics for attacks based on model-based convex optimization and generative neural networks respectively. We open-source a wave propagation and camera simulator needed for end-to-end optimization, the training software, and a library for interfacing with the camera., Comment: 30 pages, 27 figures, under review, for code, see https://github.com/ebezzam/LenslessClassification. arXiv admin note: substantial text overlap with arXiv:2206.01429

Object detectors trained with weak annotations are affordable alternatives to fully-supervised counterparts. However, there is still a significant performance gap between them. We propose to narrow this gap by fine-tuning a base pre-trained weakly-supervised detector with a few fully-annotated samples automatically selected from the training set using ``box-in-box'' (BiB), a novel active learning strategy designed specifically to address the well-documented failure modes of weakly-supervised detectors. Experiments on the VOC07 and COCO benchmarks show that BiB outperforms other active learning techniques and significantly improves the base weakly-supervised detector's performance with only a few fully-annotated images per class. BiB reaches 97% of the performance of fully-supervised Fast RCNN with only 10% of fully-annotated images on VOC07. On COCO, using on average 10 fully-annotated images per class, or equivalently 1% of the training set, BiB also reduces the performance gap (in AP) between the weakly-supervised detector and the fully-supervised Fast RCNN by over 70%, showing a good trade-off between performance and data efficiency. Our code is publicly available at https://github.com/huyvvo/BiB., Comment: Accepted to European Conference on Computer Vision (ECCV) 2022. Contains 27 pages, 9 tables and 6 figures

In this paper, we analyse propagating fronts in the context of hyperbolic theories of dissipative processes. These can be considered as a natural alternative to the more classical parabolic models. Emphasis is given toward the numerical computation of the invasion velocity., Comment: 30 pages, 5 figures, 4 tables

By replacing the lens with a thin optical element, lensless imaging enables new applications and solutions beyond those supported by traditional camera design and post-processing, e.g. compact and lightweight form factors and visual privacy. The latter arises from the highly multiplexed measurements of lensless cameras, which require knowledge of the imaging system to recover a recognizable image. In this work, we exploit this unique multiplexing property: casting the optics as an encoder that produces learned embeddings directly at the camera sensor. We do so in the context of image classification, where we jointly optimize the encoder's parameters and those of an image classifier in an end-to-end fashion. Our experiments show that jointly learning the lensless optical encoder and the digital processing allows for lower resolution embeddings at the sensor, and hence better privacy as it is much harder to recover meaningful images from these measurements. Additional experiments show that such an optimization allows for lensless measurements that are more robust to typical real-world image transformations. While this work focuses on classification, the proposed programmable lensless camera and end-to-end optimization can be applied to other computational imaging tasks., Comment: 29 pages, 23 figures, under review

Lensless imaging seeks to replace/remove the lens in a conventional imaging system. The earliest cameras were in fact lensless, relying on long exposure times to form images on the other end of a small aperture in a darkened room/container (camera obscura). The introduction of a lens allowed for more light throughput and therefore shorter exposure times, while retaining sharp focus. The incorporation of digital sensors readily enabled the use of computational imaging techniques to post-process and enhance raw images (e.g. via deblurring, inpainting, denoising, sharpening). Recently, imaging scientists have started leveraging computational imaging as an integral part of lensless imaging systems, allowing them to form viewable images from the highly multiplexed raw measurements of lensless cameras (see [5] and references therein for a comprehensive treatment of lensless imaging). This represents a real paradigm shift in camera system design as there is more flexibility to cater the hardware to the application at hand (e.g. lightweight or flat designs). This increased flexibility comes however at the price of a more demanding post-processing of the raw digital recordings and a tighter integration of sensing and computation, often difficult to achieve in practice due to inefficient interactions between the various communities of scientists involved. With LenslessPiCam, we provide an easily accessible hardware and software framework to enable researchers, hobbyists, and students to implement and explore practical and computational aspects of lensless imaging. We also provide detailed guides and exercises so that LenslessPiCam can be used as an educational resource, and point to results from our graduate-level signal processing course.

Nous nous int\'eressons \`a la reconstruction parcimonieuse d'images \`a l'aide du probl\`eme d'optimisation r\'egularis\'e LASSO. Dans de nombreuses applications pratiques, les grandes dimensions des objets \`a reconstruire limitent, voire emp\^echent, l'utilisation des m\'ethodes de r\'esolution proximales classiques. C'est le cas par exemple en radioastronomie. Nous d\'etaillons dans cet article le fonctionnement de l'algorithme \textit{Frank-Wolfe Polyatomique}, sp\'ecialement d\'evelopp\'e pour r\'esoudre le probl\`eme LASSO dans ces contextes exigeants. Nous d\'emontrons sa sup\'eriorit\'e par rapport aux m\'ethodes proximales dans des situations en grande dimension avec des mesures de Fourier, lors de la r\'esolution de probl\`emes simul\'es inspir\'es de la radio-interf\'erom\'etrie. -- We consider the problem of recovering sparse images by means of the penalised optimisation problem LASSO. For various practical applications, it is impossible to rely on the proximal solvers commonly used for that purpose due to the size of the objects to recover, as it is the case for radio astronomy. In this article we explain the mechanisms of the \textit{Polyatomic Frank-Wolfe algorithm}, specifically designed to minimise the LASSO problem in such challenging contexts. We demonstrate in simulated problems inspired from radio-interferometry the preeminence of this algorithm over the proximal methods for high dimensional images with Fourier measurements., Comment: in French

This work investigates learning pixel-wise semantic image segmentation in urban scenes without any manual annotation, just from the raw non-curated data collected by cars which, equipped with cameras and LiDAR sensors, drive around a city. Our contributions are threefold. First, we propose a novel method for cross-modal unsupervised learning of semantic image segmentation by leveraging synchronized LiDAR and image data. The key ingredient of our method is the use of an object proposal module that analyzes the LiDAR point cloud to obtain proposals for spatially consistent objects. Second, we show that these 3D object proposals can be aligned with the input images and reliably clustered into semantically meaningful pseudo-classes. Finally, we develop a cross-modal distillation approach that leverages image data partially annotated with the resulting pseudo-classes to train a transformer-based model for image semantic segmentation. We show the generalization capabilities of our method by testing on four different testing datasets (Cityscapes, Dark Zurich, Nighttime Driving and ACDC) without any finetuning, and demonstrate significant improvements compared to the current state of the art on this problem. See project webpage https://vobecant.github.io/DriveAndSegment/ for the code and more., Comment: v2: improved quality of images. See the project webpage https://vobecant.github.io/DriveAndSegment/ for the code and more

This work reports on the measured performance of an Aluminum Scandium Nitride (AlScN) Two-Dimensional Resonant Rods resonator (2DRR), fabricated by using a Sc-doping concentration of 24%, characterized by a low off-resonance impedance (~25 Ohm) and exhibiting a record electromechanical coupling coefficient (kt2) of 23.9% for AlScN resonators. In order to achieve such performance, we identified and relied on optimized deposition and etching processes for highly-doped AlScN films, aiming at achieving high crystalline quality, low density of abnormally oriented grains in the 2DRR's active region and sharp lateral sidewalls. Also, the 2DRR's unit-cell has been acoustically engineered to maximize the piezo-generated mechanical energy within each rod and to ensure a low transduction of spurious modes around resonance. Due to its unprecedented kt2, the reported 2DRR opens exciting scenarios towards the development of next generation monolithic integrated radio-frequency (RF) filtering components. In fact, we show that 5th-order 2DRR-based ladder filters with fractional bandwidths (BW) of ~11%, insertion-loss (I.L) values of ~2.5 dB and with >30 dB out-of-band rejections can now be envisioned, paving an unprecedented path towards the development of ultra-wide band (UWB) filters for next-generation Super-High-Frequency (SHF) radio front-ends., Comment: 10 pages, 11 figures, 5 tables, submitted to JMEMS

In this work we provide analytic and numerical solutions for the Bjorken flow, a standard benchmark in relativistic hydrodynamics providing a simple model for the bulk evolution of matter created in collisions between heavy nuclei. We consider relativistic gases of both massive and massless particles, working in a ( 2 + 1 ) and ( 3 + 1 ) Minkowski space-time coordinate system. The numerical results from a recently developed lattice kinetic scheme show excellent agreement with the analytic solutions.

We present a new family of relativistic lattice kinetic schemes for the efficient simulation of relativistic flows in both strongly-interacting (fluid) and weakly-interacting (rarefied gas) regimes. The method can also deal with both massless and massive particles, thereby encompassing ultra-relativistic and mildly-relativistic regimes alike. The computational performance of the method for the simulation of relativistic flows across the aforementioned regimes is discussed in detail, along with prospects of future applications for Quark-Gluon Plasma, electron flows in graphene and systems in astrophysical contexts., Comment: 34 pages, 6 figures

We propose nonparametric estimators for the second-order central moments of possibly anisotropic spherical random fields, within a functional data analysis context. We consider a measurement framework where each random field among an identically distributed collection of spherical random fields is sampled at a few random directions, possibly subject to measurement error. The collection of random fields could be i.i.d. or serially dependent. Though similar setups have already been explored for random functions defined on the unit interval, the nonparametric estimators proposed in the literature often rely on local polynomials, which do not readily extend to the (product) spherical setting. We therefore formulate our estimation procedure as a variational problem involving a generalized Tikhonov regularization term. The latter favours smooth covariance/autocovariance functions, where the smoothness is specified by means of suitable Sobolev-like pseudo-differential operators. Using the machinery of reproducing kernel Hilbert spaces, we establish representer theorems that fully characterize the form of our estimators. We determine their uniform rates of convergence as the number of random fields diverges, both for the dense (increasing number of spatial samples) and sparse (bounded number of spatial samples) regimes. We moreover demonstrate the computational feasibility and practical merits of our estimation procedure in a simulation setting, assuming a fixed number of samples per random field. Our numerical estimation procedure leverages the sparsity and second-order Kronecker structure of our setup to reduce the computational and memory requirements by approximately three orders of magnitude compared to a naive implementation would require.

We propose a fast and scalable Polyatomic Frank-Wolfe (P-FW) algorithm for the resolution of high-dimensional LASSO regression problems. The latter improves upon traditional Frank-Wolfe methods by considering generalized greedy steps with polyatomic (i.e. linear combinations of multiple atoms) update directions, hence allowing for a more efficient exploration of the search space. To preserve sparsity of the intermediate iterates, we re-optimize the LASSO problem over the set of selected atoms at each iteration. For efficiency reasons, the accuracy of this re-optimization step is relatively low for early iterations and gradually increases with the iteration count. We provide convergence guarantees for our algorithm and validate it in simulated compressed sensing setups. Our experiments reveal that P-FW outperforms state-of-the-art methods in terms of runtime, both for FW methods and optimal first-order proximal gradient methods such as the Fast Iterative Soft-Thresholding Algorithm (FISTA)., Comment: Published in IEEE Signal Processing Letters

Fourier transforms are an often necessary component in many computational tasks, and can be computed efficiently through the fast Fourier transform (FFT) algorithm. However, many applications involve an underlying continuous signal, and a more natural choice would be to work with e.g. the Fourier series (FS) coefficients in order to avoid the additional overhead of translating between the analog and discrete domains. Unfortunately, there exists very little literature and tools for the manipulation of FS coefficients from discrete samples. This paper introduces a Python library called pyFFS for efficient FS coefficient computation, convolution, and interpolation. While the libraries SciPy and NumPy provide efficient functionality for discrete Fourier transform coefficients via the FFT algorithm, pyFFS addresses the computation of FS coefficients through what we call the fast Fourier series (FFS). Moreover, pyFFS includes an FS interpolation method based on the chirp Z-transform that can make it more than an order of magnitude faster than the SciPy equivalent when one wishes to perform interpolation. GPU support through the CuPy library allows for further acceleration, e.g. an order of magnitude faster for computing the 2-D FS coefficients of 1000 x 1000 samples and nearly two orders of magnitude faster for 2-D interpolation. As an application, we discuss the use of pyFFS in Fourier optics. pyFFS is available as an open source package at https://github.com/imagingofthings/pyFFS, with documentation at https://pyffs.readthedocs.io., Comment: SIAM Journal on Scientific Computing, Volume 44, Issue 4 (2022)

Localizing objects in image collections without supervision can help to avoid expensive annotation campaigns. We propose a simple approach to this problem, that leverages the activation features of a vision transformer pre-trained in a self-supervised manner. Our method, LOST, does not require any external object proposal nor any exploration of the image collection; it operates on a single image. Yet, we outperform state-of-the-art object discovery methods by up to 8 CorLoc points on PASCAL VOC 2012. We also show that training a class-agnostic detector on the discovered objects boosts results by another 7 points. Moreover, we show promising results on the unsupervised object discovery task. The code to reproduce our results can be found at https://github.com/valeoai/LOST.

The Gatenby-Gawlinski model for cancer invasion is object of analysis in order to investigate the mathematical framework behind the model working by means of suitable reductions. We perform numerical simulations to study the sharpness/smoothness of the traveling fronts starting from a brief overview about the full model and proceed by examining the case of a two-equations-based and one-equation-based reduction. We exploit a numerical strategy depending on a finite volume approximation and employ a space-averaged wave speed estimate to quantitatively approach the traveling waves phenomenon. Concerning the one equation-based model, we propose a reduction framed within the degenerate reaction-diffusion equations field, which proves to be effective in order to qualitatively recover the typical trends arising from the Gatenby-Gawlinski model. Finally, we carry out some numerical tests in a specific case where the analytical solution is available., Comment: 21 pages, 13 figures

In this article, we propose a mathematical model -- based on a cellular automaton -- for the redistribution of patients within a network of hospitals with limited available resources, in order to reduce the risks of a local/global collapse of the healthcare system. We attempt at developing a conceptual tool to support making rational decisions relevant to the optimisation of the allocation of patients into accessible medical facilities. The strategy is based on a version of the Abelian Sandpile model for the Self-Organised Criticality, with the idea of testing the paradigm for the management of patients among the COVID-19 hospitals in Italian regions. In particular, we compare the novel proposal to the standard management of connections between hospitals, showing a number of advantages at a local and global level, by means of a reliable indicator function introduced for measuring the effectiveness of the allocation strategies., Comment: 24 pages, 8 figures

Stochastic Maximum Likelihood (SML) is a popular direction of arrival (DOA) estimation technique in array signal processing. It is a parametric method that jointly estimates signal and instrument noise by maximum likelihood, achieving excellent statistical performance. Some drawbacks are the computational overhead as well as the limitation to a point-source data model with fewer sources than sensors. In this work, we propose a Sieved Maximum Likelihood (SiML) method. It uses a general functional data model, allowing an unrestricted number of arbitrarily-shaped sources to be recovered. To this end, we leverage functional analysis tools and express the data in terms of an infinite-dimensional sampling operator acting on a Gaussian random function. We show that SiML is computationally more efficient than traditional SML, resilient to noise, and results in much better accuracy than spectral-based methods., Comment: 5 pages, 2 figures. Published in ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), scheduled for 6-11 June 2021 in Toronto, Ontario, Canada