Search

Accurately detecting 3D objects from monocular images in dynamic roadside scenarios remains a challenging problem due to varying camera perspectives and unpredictable scene conditions. This paper introduces a two-stage training strategy to address these challenges. Our approach initially trains a model on the large-scale synthetic dataset, RoadSense3D, which offers a diverse range of scenarios for robust feature learning. Subsequently, we fine-tune the model on a combination of real-world datasets to enhance its adaptability to practical conditions. Experimental results of the Cube R-CNN model on challenging public benchmarks show a remarkable improvement in detection performance, with a mean average precision rising from 0.26 to 12.76 on the TUM Traffic A9 Highway dataset and from 2.09 to 6.60 on the DAIR-V2X-I dataset when performing transfer learning. Code, data, and qualitative video results are available on the project website: https://roadsense3d.github.io., Comment: 18 pages. Accepted for ECVA European Conference on Computer Vision 2024 (ECCV'24)

In this paper, we investigate the reduction process of a contact Lagrangian system whose Lagrangian is invariant under a group of symmetries. We give explicit coordinate expressions of the resulting reduced differential equations, the so-called Lagrange-Poincare-Herglotz equations. Our framework relied on the associated Herglotz vector field and its projected vector field, and the use of well-chosen quasi-velocities. Some examples are also discussed., Comment: 31 pages

This letter illustrates the opinion of the molecular dynamics (MD) community on the need to adopt a new FAIR paradigm for the use of molecular simulations. It highlights the necessity of a collaborative effort to create, establish, and sustain a database that allows findability, accessibility, interoperability, and reusability of molecular dynamics simulation data. Such a development would democratize the field and significantly improve the impact of MD simulations on life science research. This will transform our working paradigm, pushing the field to a new frontier. We invite you to support our initiative at the MDDB community (https://mddbr.eu/community/)

Noise resilience of quantum information processing is a crucial precondition to reach the fault-tolerance threshold. While resilience to many types of noise can be achieved through suitable control schemes, resilience to amplitude noise seems to be elusive within the common harmonic approximation for the bus mode of trapped ions. We show that weak an-harmonicities admit control schemes that achieve amplitude noise-resilience consistent with state-of-the-art experimental requirements, and that the required an-harmonicities can be achieved with current standards of micro-structured traps or even the intrinsically an-harmonic Coulomb interaction.

We show that combining ideas from the fields of quantum invariants and of optimal control can be used to design quantum control of quantum systems without explicit reference to quantum states. The scaling in numerical effort of the resultant approach is given by commutation relations of system operators, and it can be polynomial in the number of subsystems despite the general quantum mechanical exponential scaling of the Hilbert space. As explicit applications, we discuss state preparation and quantum simulation with Hamiltonians including three-body and many-body interactions with spin chains of up to 50 constituents, and the perspective of use for topologically protected quantum information processing., Comment: 14 pages, 4 figures

We discuss the conical region of convergence of exponential and asymptotically polynomial form factors and their integral representations. Then, we calculate the spectral representation of the propagator of nonlocal theories with entire form factors, in particular, of the above type. The spectral density is positive-definite and exhibits the same spectrum as the local theory. We also find that the piece of the propagator corresponding to the time-ordered two-point correlation function admits a generalization of the K\"all\'en-Lehmann representation with a standard momentum dependence and a spectral density differing from the local one only in the presence of interactions. These results are in agreement with what already known about the free theory after a field redefinition and about perturbative unitarity of the interacting theory. The spectral and K\"all\'en-Lehmann representations have the same standard local limit, which is recovered smoothly when sending the fundamental length scale $\ell_*$ in the form factor to zero., Comment: 1+31 pages, 1 figure. v2: definition of nonlocal operators and local limit clarified

We present an entangling gate scheme for trapped-ion chains that achieves high-fidelity operations with excited motional states despite multiple error sources. Our approach incorporates all relevant motional modes and exhibits enhanced robustness against both motional heating effects and detuning errors, critical features for building robust and scalable trapped-ion quantum computers., Comment: 11 pages, 4 figures

Emotion classification through EEG signals plays a significant role in psychology, neuroscience, and human-computer interaction. This paper addresses the challenge of mapping human emotions using EEG data in the Mapping Human Emotions through EEG Signals FG24 competition. Subjects mimic the facial expressions of an avatar, displaying fear, joy, anger, sadness, disgust, and surprise in a VR setting. EEG data is captured using a multi-channel sensor system to discern brain activity patterns. We propose a novel two-stream neural network employing a Bi-Hemispheric approach for emotion inference, surpassing baseline methods and enhancing emotion recognition accuracy. Additionally, we conduct a temporal analysis revealing that specific signal intervals at the beginning and end of the emotion stimulus sequence contribute significantly to improve accuracy. Leveraging insights gained from this temporal analysis, our approach offers enhanced performance in capturing subtle variations in the states of emotions., Comment: Second place award at the Brain Responses to Emotional Avatars Challenge held by the 18th IEEE International Conference on Automatic Face and Gesture Recognition(FG2024)

Ribonucleic acids (RNA) are unique in that they can store genetic information, replicate and perform catalysis. Importantly, RNA molecules are highly dynamic, and thus determining the ensemble of conformations that they populate is crucial not only to elucidate their biological functions, but also for their potential use as therapeutic targets. Computational and experimental techniques provide complementary views on RNA dynamics, and their integration is fundamental to improve the accuracy of computations and the resolution of experiments. Recent exciting developments in this field, were discussed at the CECAM workshop ``RNA dynamics from experimental and computational approaches'', in Paris, June 26-28, 2023. This report outlines key `take-home' messages that emerged during this workshop from the presentations and discussions.

We introduce the Lorentzian path integral of nonlocal quantum gravity. After introducing the functional measure, the Faddeev-Popov sector and the field correlators, we move to perturbation theory and describe Efimov analytic continuation of scattering amplitudes to Euclidean momenta and back to Lorentzian. We show that the conformal instability problem in the Euclidean path integral is solved by suitable gauge choices at the perturbative level. The three examples of Einstein gravity, Stelle gravity and nonlocal quantum gravity are given., Comment: 1+27 pages. v2: sections 2.3, 2.5 and 4 revised and expanded, Efimov analytic continuation and relation between Lorentzian and Euclidean path integrals clarified, conclusions on conformal instability strengthened (section 5), references added

The Open Databases Integration for Materials Design (OPTIMADE) application programming interface (API) empowers users with holistic access to a growing federation of databases, enhancing the accessibility and discoverability of materials and chemical data. Since the first release of the OPTIMADE specification (v1.0), the API has undergone significant development, leading to the upcoming v1.2 release, and has underpinned multiple scientific studies. In this work, we highlight the latest features of the API format, accompanying software tools, and provide an update on the implementation of OPTIMADE in contributing materials databases. We end by providing several use cases that demonstrate the utility of the OPTIMADE API in materials research that continue to drive its ongoing development.

An algorithm is proposed that enables the calculation of pre-neutron mass and energy distributions of fission fragments from uranium-235, induced by thermal neutrons, utilizing measurements obtained through the 2E technique. This algorithm facilitates the inference of the curve representing the average multiplicity of prompt neutrons relative to the pre-neutron mass of the fission fragments., Comment: 8 pages, 3 figures

Re-identifying participants in ultra-distance running competitions can be daunting due to the extensive distances and constantly changing terrain. To overcome these challenges, computer vision techniques have been developed to analyze runners' faces, numbers on their bibs, and clothing. However, our study presents a novel gait-based approach for runners' re-identification (re-ID) by leveraging various pre-trained human action recognition (HAR) models and loss functions. Our results show that this approach provides promising results for re-identifying runners in ultra-distance competitions. Furthermore, we investigate the significance of distinct human body movements when athletes are approaching their endurance limits and their potential impact on re-ID accuracy. Our study examines how the recognition of a runner's gait is affected by a competition's critical point (CP), defined as a moment of severe fatigue and the point where the finish line comes into view, just a few kilometers away from this location. We aim to determine how this CP can improve the accuracy of athlete re-ID. Our experimental results demonstrate that gait recognition can be significantly enhanced (up to a 9% increase in mAP) as athletes approach this point. This highlights the potential of utilizing gait recognition in real-world scenarios, such as ultra-distance competitions or long-duration surveillance tasks., Comment: Accepted at 7th International Joint Conference on Biometrics (IJCB 2023)

The gravitational induced interference is here studied in the framework of Teleparallel Gravity. We derive the gravitational phase difference and we apply the result to the case of a Kerr spacetime. Afterwards, we compute the fringe shifts in an interference experiment of particles and discuss how to increase their values by changing the given parameters that include: the area in between the paths, the energy of the particles, the distance from the black hole, the mass and the spin of the black hole. It turns out that it is more difficult to detect the fringe shifts for massless particles than for massive particles. As a further application, we show how the mass of the black hole and its angular momentum can be obtained from the measurement of the fringe shifts. Finally, we compare the phase difference derived in Teleparallel Gravity with a previous work in General Relativity., Comment: 24 pages, 2 figures, accepted for publication in Physical Review D, see https://journals.aps.org/prd/accepted/0b07dQ6ePa71133d22891ad3d1b8589b88a77bffc

In this paper we carry out a detailed analysis of the static spherically symmetric solutions of a sixth-derivative Lee--Wick gravity model in the effective delta source approximation. Previous studies of these solutions have only considered the particular case in which the real and the imaginary part of the Lee--Wick mass $\mu=a +i b$ are equal. However, as we show here, the solutions exhibit an interesting structure when the full parameter space is considered, owing to the oscillations of the metric that depend on the ratio $b/a$. Such oscillations can generate a rich structure of horizons, a sequence of mass gaps and the existence of multiple regimes for black hole sizes (horizon position gaps). In what concerns the thermodynamics of these objects, the oscillation of the Hawking temperature determines the presence of multiple mass scales for the remnants of the evaporation process and may permit the existence of cold black holes with zero Hawking temperature~$T$ and quasi-stable intermediate configurations with $T \approx 0$ and a long evaporation lifetime. For the sake of generality, we consider two families of solutions, one with a trivial shift function and the other with a non-trivial one (dirty black hole). The latter solution has the advantage of reproducing the modified Newtonian-limit metric of Lee--Wick gravity for small and large values of~$r$., Comment: 23 pages, 10 figures, 4 tables

In this work we study static spherically symmetric solutions of effective field equations related to local and nonlocal higher-derivative gravity models, based on their associated effective delta sources. This procedure has been applied to generate modifications of the Schwarzschild geometry in several contexts (e.g., modified gravity, string theory, noncommutative geometry, generalized uncertainty principle scenarios), but a general analysis of the possible equations of state and their influence on the solutions was still lacking. Here, we aim to fill this gap in the literature and investigate whether these metrics might be able to reproduce features of the solutions of higher-derivative gravity models. In particular, we present an equation of state such that the solution matches the Newtonian-limit one in both regimes of large and small $r$. A significant part of the work is dedicated to studying the curvature regularity of the solutions and the comparison with the linearized solutions. Explicit metrics are presented for effective sources originating from local and nonlocal models. The results obtained here might be regarded as possible links between the previous research on linearized higher-derivative gravity and the solutions of the nonlinear complete field equations, which remain unknown at the moment., Comment: 33 pages, 8 figures. Extended discussion, references added; matches the published version

A Monte Carlo simulation algorithm to investigate the measurement of the average prompt neutron multiplicity as a function of pre-neutron mass, for fragments from the spontaneous fission of 252Cf is presented. The input data consist of experimental measurements of the kinetic energy and mass distributions obtained by Gook et al. and the values calculated using the FIFRELIN model by Piau et al. We analyze the output curve obtained by simulation of the 2E technique, which should ideally match the theoretical curve. However, we find that the simulated as experiential curve exhibits a maximum value at A=122, close to mass symmetry, while the pre-neutron curve has a maximum at A=118. Additionally, we observe that que the simulated as pre neutron values are higher than the theoretical values for A<90 and A>162, respectively. We attribute this discrepancy to inaccuracies in the relationship between provisional mass and the pre-neutron mass used in the 2E technique for data processing in each fission event., Comment: 6 pages, 3 figures

In this paper, we introduce a geometric description of contact Lagrangian and Hamiltonian systems on Lie algebroids in the framework of contact geometry, using the theory of prolongations. We discuss the relation between Lagrangian and Hamiltonian settings through a convenient notion of Legendre transformation. We also discuss the Hamilton-Jacobi problem in this framework and introduce the notion of a Legendrian Lie subalgebroid of a contact Lie algebroid., Comment: 36 pages. arXiv admin note: text overlap with arXiv:2204.11537 by other authors

We present a transfer learning analysis on a sporting environment of the expanded 3D (X3D) neural networks. Inspired by action quality assessment methods in the literature, our method uses an action recognition network to estimate athletes' cumulative race time (CRT) during an ultra-distance competition. We evaluate the performance considering the X3D, a family of action recognition networks that expand a small 2D image classification architecture along multiple network axes, including space, time, width, and depth. We demonstrate that the resulting neural network can provide remarkable performance for short input footage, with a mean absolute error of 12 minutes and a half when estimating the CRT for runners who have been active from 8 to 20 hours. Our most significant discovery is that X3D achieves state-of-the-art performance while requiring almost seven times less memory to achieve better precision than previous work., Comment: Accepted to the 18th International Conference on Machine Vision Applications (MVA 2023)

We extensively study the ultraviolet quantum properties of a nonlocal action for gravity nonminimally coupled to matter. The theory unifies matter and gravity in an action principle such that all the classical solutions of Einstein's theory coupled to matter are also solutions of the nonlocal theory. At the quantum level, we show that the theory is power-counting super-renormalizable in even dimensions and finite in odd dimensions. A simple extension of the model compatible with the above properties is finite also in even dimensions., Comment: 52 pages. v2: abstract simplified, terminology denoting the extremals changed, some parts clarified, new section 6.4 added, conclusions unchanged. Accepted for publication in JHEP

We extend the Jacobi structure from $TQ\times \mathbb{R}$ and $T^{*}Q \times \mathbb{R}$ to $A\times \mathbb{R}$ and $A^{*}\times \mathbb{R}$, respectively, where $A$ is a Lie algebroid and $A^{*}$ carries the associated Poisson structure. We see that $A^*\times \mathbb{R}$ possesses a natural Jacobi structure from where we are able to model dissipative mechanical systems, generalizing previous models on $TQ\times \mathbb{R}$ and $\mathfrak{g}\times \mathbb{R}$., Comment: 16 pages

In various types of many-particle systems, bidispersity is frequently used to avoid spontaneous ordering in particle configuration. In this study, the relation between bidispersity and disorder degree of particle configuration is investigated. By using magnetic dipole-dipole interaction, magnet particles are dispersed in a two-dimensional cell without physical contact between them. In this magnetic system, bidispersity is introduced by mixing large and small magnets. Then, the particle system is compressed to produce a uniform particle configuration. The compressed particle configuration is analyzed by using Voronoi tessellation for evaluating the disorder degree which strongly depends on bidispersity. Specifically, standard deviation and skewness of the Voronoi cell area distribution are measured. As a result, we find that the peak of standard deviation is observed when the numbers of large and small particles are almost identical. Although the skewness shows non-monotonic behavior, zero skewness state (symmetric distribution) can be achieved when the numbers of large and small particles are identical. In this ideally random (disordered) state, the ratio between pentagonal, hexagonal, and heptagonal Voronoi cells become roughly identical, while hexagons are dominant in monodisperse (ordered) condition. The relation between Voronoi cell analysis and the global bond orientational order parameter is also discussed., Comment: 9 pages, 10 figures