Search

Single-shot hyperspectral wavefront sensing is essential for applications like spatio-spectral coupling metrology in high power laser or fast material dispersion imaging. Under broadband illumination, traditional wavefront sensors assume an achromatic wavefront, which makes them unsuitable. We introduce a hyperspectral wavefront sensing scheme based on the Hartmann wavefront sensing principles, employing a multicore fiber as a modified Hartmann mask to overcome these limitations. Our system leverages the angular memory effect and spectral decorrelation from the multicore fiber, encoding wavefront gradients into displacements and the spectral information into uncorrelated patterns. This method retains the simplicity, compactness, and single-shot capability of conventional wavefront sensors, with only a slight increase in computational complexity. It also allows a tunable trade-off between spatial and spectral resolution. We demonstrate its efficacy for recording the hyperspectral wavefront cube from single-pulse acquisitions at the Apollon multi-PW laser facility, and for performing multispectral microscopic imaging of dispersive phase objects.

Multi-agent combinatorial optimization problems such as routing and scheduling have great practical relevance but present challenges due to their NP-hard combinatorial nature, hard constraints on the number of possible agents, and hard-to-optimize objective functions. This paper introduces PARCO (Parallel AutoRegressive Combinatorial Optimization), a novel approach that learns fast surrogate solvers for multi-agent combinatorial problems with reinforcement learning by employing parallel autoregressive decoding. We propose a model with a Multiple Pointer Mechanism to efficiently decode multiple decisions simultaneously by different agents, enhanced by a Priority-based Conflict Handling scheme. Moreover, we design specialized Communication Layers that enable effective agent collaboration, thus enriching decision-making. We evaluate PARCO in representative multi-agent combinatorial problems in routing and scheduling and demonstrate that our learned solvers offer competitive results against both classical and neural baselines in terms of both solution quality and speed. We make our code openly available at https://github.com/ai4co/parco.

This paper introduces RouteFinder, a comprehensive foundation model framework to tackle different Vehicle Routing Problem (VRP) variants. Our core idea is that a foundation model for VRPs should be able to represent variants by treating each as a subset of a generalized problem equipped with different attributes. We propose a unified VRP environment capable of efficiently handling any attribute combination. The RouteFinder model leverages a modern transformer-based encoder and global attribute embeddings to improve task representation. Additionally, we introduce two reinforcement learning techniques to enhance multi-task performance: mixed batch training, which enables training on different variants at once, and multi-variant reward normalization to balance different reward scales. Finally, we propose efficient adapter layers that enable fine-tuning for new variants with unseen attributes. Extensive experiments on 24 VRP variants show RouteFinder achieves competitive results. Our code is openly available at https://github.com/ai4co/routefinder., Comment: A version of this work has been presented as an Oral at the ICML 2024 FM-Wild Workshop

Traffic incidents involving vulnerable road users (VRUs) constitute a significant proportion of global road accidents. Advances in traffic communication ecosystems, coupled with sophisticated signal processing and machine learning techniques, have facilitated the utilization of data from diverse sensors. Despite these advancements and the availability of extensive datasets, substantial progress is required to mitigate traffic casualties. This paper provides a comprehensive survey of state-of-the-art technologies and methodologies to enhance the safety of VRUs. The study delves into the communication networks between vehicles and VRUs, emphasizing the integration of advanced sensors and the availability of relevant datasets. It explores preprocessing techniques and data fusion methods to enhance sensor data quality. Furthermore, our study assesses critical simulation environments essential for developing and testing VRU safety systems. Our research also highlights recent advances in VRU detection and classification algorithms, addressing challenges such as variable environmental conditions. Additionally, we cover cutting-edge research in predicting VRU intentions and behaviors, which is crucial for proactive collision avoidance strategies. Through this survey, we aim to provide a comprehensive understanding of the current landscape of VRU safety technologies, identifying areas of progress and areas needing further research and development., Comment: 57 pages, 18 tables, 8 figures, citing 339 (up-to-date) papers, preprint submitted to Expert Systems with Applications (Elsevier)

Multi-Agent Reinforcement Learning (MARL) based Multi-Agent Path Finding (MAPF) has recently gained attention due to its efficiency and scalability. Several MARL-MAPF methods choose to use communication to enrich the information one agent can perceive. However, existing works still struggle in structured environments with high obstacle density and a high number of agents. To further improve the performance of the communication-based MARL-MAPF solvers, we propose a new method, Ensembling Prioritized Hybrid Policies (EPH). We first propose a selective communication block to gather richer information for better agent coordination within multi-agent environments and train the model with a Q learning-based algorithm. We further introduce three advanced inference strategies aimed at bolstering performance during the execution phase. First, we hybridize the neural policy with single-agent expert guidance for navigating conflict-free zones. Secondly, we propose Q value-based methods for prioritized resolution of conflicts as well as deadlock situations. Finally, we introduce a robust ensemble method that can efficiently collect the best out of multiple possible solutions. We empirically evaluate EPH in complex multi-agent environments and demonstrate competitive performance against state-of-the-art neural methods for MAPF. We open-source our code at https://github.com/ai4co/eph-mapf., Comment: Accepted to 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2024)

Large-scale multi-agent pathfinding (MAPF) presents significant challenges in several areas. As systems grow in complexity with a multitude of autonomous agents operating simultaneously, efficient and collision-free coordination becomes paramount. Traditional algorithms often fall short in scalability, especially in intricate scenarios. Reinforcement Learning (RL) has shown potential to address the intricacies of MAPF; however, it has also been shown to struggle with scalability, demanding intricate implementation, lengthy training, and often exhibiting unstable convergence, limiting its practical application. In this paper, we introduce Heuristics-Informed Multi-Agent Pathfinding (HiMAP), a novel scalable approach that employs imitation learning with heuristic guidance in a decentralized manner. We train on small-scale instances using a heuristic policy as a teacher that maps each single agent observation information to an action probability distribution. During pathfinding, we adopt several inference techniques to improve performance. With a simple training scheme and implementation, HiMAP demonstrates competitive results in terms of success rate and scalability in the field of imitation-learning-only MAPF, showing the potential of imitation-learning-only MAPF equipped with inference techniques., Comment: Accepted as Extended Abstract in Proc. of the 23rd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2024)

We present our optical photometric observations of the 2022 eruption of the recurrent nova U Scorpii (U Sco) using 49,152 data points over 70 d following the optical peak. We have also analyzed its soft X-ray (0.3--1 keV) light curve by the Neil Gehrels Swift Observatory. During the 2022 eruption, the optical plateau stage started 13.8--15.0 d and ended 23.8--25.0 d after the optical peak. The soft X-ray stage started 14.6--15.3 d and ended 38.7--39.5 d after the optical peak. Both stages started later and had shorter durations, and the soft X-ray light curve peaked earlier and was less luminous compared to those during the U Sco 2010 eruption. These points suggest that there were differences in the envelope mass between the different cycles of the nova eruption. Furthermore, we have analyzed the optical eclipses during the 2022 eruption. The primary eclipse was first observed 10.4--11.6 d after the optical peak, earlier than the beginning of the optical plateau stage. This sequence of events can be explained by the receding ejecta photosphere associated with the expanding nova ejecta. We have determined the ingress and egress phases of the primary eclipses and estimated the outer radius of the optical light source centered at the white dwarf (WD). During the optical plateau stage, the source radius remained $\sim$1.2 times larger than the Roche volume radius of the primary WD, being close to the L1 point. When the optical plateau stage ended, the source radius drastically shrank to the tidal truncation radius within a few orbital periods. This previously unresolved phenomenon can be interpreted as a structural change in U Sco where the temporarily expanded accretion disk due to the nova wind returned to a steady state., Comment: 16 pages, 7 figures, 7 tables, accepted for publication in PASJ; doi:10.1093/pasj/psae010

The omnipresence of NP-hard combinatorial optimization problems (COPs) compels domain experts to engage in trial-and-error heuristic design. The long-standing endeavor of design automation has gained new momentum with the rise of large language models (LLMs). This paper introduces Language Hyper-Heuristics (LHHs), an emerging variant of Hyper-Heuristics that leverages LLMs for heuristic generation, featuring minimal manual intervention and open-ended heuristic spaces. To empower LHHs, we present Reflective Evolution (ReEvo), a novel integration of evolutionary search for efficiently exploring the heuristic space, and LLM reflections to provide verbal gradients within the space. Across five heterogeneous algorithmic types, six different COPs, and both white-box and black-box views of COPs, ReEvo yields state-of-the-art and competitive meta-heuristics, evolutionary algorithms, heuristics, and neural solvers, while being more sample-efficient than prior LHHs., Comment: Accepted at NeurIPS 2024

While complex simulations of physical systems have been widely used in engineering and scientific computing, lowering their often prohibitive computational requirements has only recently been tackled by deep learning approaches. In this paper, we present GraphSplineNets, a novel deep-learning method to speed up the forecasting of physical systems by reducing the grid size and number of iteration steps of deep surrogate models. Our method uses two differentiable orthogonal spline collocation methods to efficiently predict response at any location in time and space. Additionally, we introduce an adaptive collocation strategy in space to prioritize sampling from the most important regions. GraphSplineNets improve the accuracy-speedup tradeoff in forecasting various dynamical systems with increasing complexity, including the heat equation, damped wave propagation, Navier-Stokes equations, and real-world ocean currents in both regular and irregular domains., Comment: Published as a conference paper in NeurIPS 2023

Superconducting quantum processors have made significant progress in size and computing potential. As a result, the practical cryogenic limitations of operating large numbers of superconducting qubits are becoming a bottleneck for further scaling. Due to the low thermal conductivity and the dense optical multiplexing capacity of telecommunications fiber, converting qubit signal processing to the optical domain using microwave-to-optics transduction would significantly relax the strain on cryogenic space and thermal budgets. Here, we demonstrate optical readout through an optical fiber of a superconducting transmon qubit connected via a coaxial cable to a fully integrated piezo-optomechanical transducer. Using a demolition readout technique, we achieve a single shot readout fidelity of 81%. Due to the small footprint (<0.15mm$^2$) and the modular fiber-based architecture, this device platform has the potential to scale towards use with thousands of qubits. Our results illustrate the potential of piezo-optomechanical transduction for low-dissipation operation of large quantum processors.

EX Lup is the archetype for the class of young stars that undergoes repeated accretion outbursts of $\sim 5$ mag at optical wavelengths and that last for months. Despite extensive monitoring that dates back 130 years, the accretion history of EX Lup remains mostly qualitative and has large uncertainties. We assess historical accretion rates of EX Lup by applying correlations between optical brightness and accretion, developed on multi-band magnitude photometry of the $\sim 2$ mag optical burst in 2022. Two distinct classes of bursts occur: major outbursts ($\Delta V\sim5$ mag) have year-long durations, are rare, reach accretion rates of $\dot{M}_{\rm acc}\sim10^{-7}~M_\odot~{\rm yr^{-1}}$ at peak, and have a total accreted mass of around 0.1 Earth masses. The characteristic bursts ($\Delta V\sim2$ mag) have durations of $\sim 2-3$ months, are more common, reach accretion rates of $\dot{M}_{\rm acc}\sim10^{-8}~M_\odot~{\rm yr^{-1}}$ at peak, and have a total accreted mass of around $10^{-3}$ Earth masses. The distribution of total accreted mass in the full set of bursts is poorly described by a power law, which suggests different driving causes behind the major outburst and characteristic bursts. The total mass accreted during two classes of bursts is around two times the masses accreted during quiescence. Our analysis of the light curves reveals a color-dependent time lag in the 2022 post-burst light curve, attributed to the presence of both hot and cool spots on the stellar surface., Comment: Accepted by ApJ; 29 pages,19 figures, 6 tables

ASASSN-22ak is a transient discovered by the All-Sky Automated Survey for Supernovae and by Gaia in 2022 January. Although this object had been in deep quiescence at least for seven years before this outburst, it has been showing relatively regular long (35-40 d) outbursts with intervals of 132-188 d after the 2022 January outburst. This "waking up" phenomenon appears similar to the very unusual (hydrogen-rich) WZ Sge star V3101 Cyg. Time-resolved photometry during the 2023 outburst detected low-amplitude (0.05 mag) superhumps with a period of 0.042876(3) d. ASASSN-22ak appears to be very similar to CRTS J112253.3-111037, which is known to have a very low mass ratio and is considered to be an object evolving close to AM CVn stars as inferred from the low hydrogen and high helium content. ASASSN-22ak is likely yet another object having an evolved core and strongly depleted hydrogen in the secondary. The case of ASASSN-22ak strengthens the idea that a considerable fraction of AM CVn stars are formed from evolved cataclysmic variables. Both ASASSN-22ak and V3101 Cyg before the initial outbursts were probably in dormant states with low quiescent viscosity or low mass-transfer rates. The current "high" states of ASASSN-22ak and V3101 Cyg may have been induced by radiation during the initial outburst or these objects are simply returning to ordinary states, either in terms of quiescent viscosity or mass-transfer rates. We also provide updated superhump period and estimated mass ratio for CRTS J112253.3-111037., Comment: 17 pages, 6 figures, VSOLJ Variable Star Bulletin No. 120

Deep reinforcement learning (RL) has recently shown significant benefits in solving combinatorial optimization (CO) problems, reducing reliance on domain expertise, and improving computational efficiency. However, the field lacks a unified benchmark for easy development and standardized comparison of algorithms across diverse CO problems. To fill this gap, we introduce RL4CO, a unified and extensive benchmark with in-depth library coverage of 23 state-of-the-art methods and more than 20 CO problems. Built on efficient software libraries and best practices in implementation, RL4CO features modularized implementation and flexible configuration of diverse RL algorithms, neural network architectures, inference techniques, and environments. RL4CO allows researchers to seamlessly navigate existing successes and develop their unique designs, facilitating the entire research process by decoupling science from heavy engineering. We also provide extensive benchmark studies to inspire new insights and future work. RL4CO has attracted numerous researchers in the community and is open-sourced at https://github.com/ai4co/rl4co., Comment: A previous version was presented as a workshop paper at the NeurIPS 2023 GLFrontiers Workshop (Oral)

We study the problem of optimizing biological sequences, e.g., proteins, DNA, and RNA, to maximize a black-box score function that is only evaluated in an offline dataset. We propose a novel solution, bootstrapped training of score-conditioned generator (BootGen) algorithm. Our algorithm repeats a two-stage process. In the first stage, our algorithm trains the biological sequence generator with rank-based weights to enhance the accuracy of sequence generation based on high scores. The subsequent stage involves bootstrapping, which augments the training dataset with self-generated data labeled by a proxy score function. Our key idea is to align the score-based generation with a proxy score function, which distills the knowledge of the proxy score function to the generator. After training, we aggregate samples from multiple bootstrapped generators and proxies to produce a diverse design. Extensive experiments show that our method outperforms competitive baselines on biological sequential design tasks. We provide reproducible source code: \href{https://github.com/kaist-silab/bootgen}{https://github.com/kaist-silab/bootgen}., Comment: NeurIPS 2023, 19 pages, 5 figures

Neuroscience research has expanded dramatically over the past 30 years by advancing standardization and tool development to support rigor and transparency. Consequently, the complexity of the data pipeline has also increased, hindering access to FAIR (Findable, Accessible, Interoperabile, and Reusable) data analysis to portions of the worldwide research community. brainlife.io was developed to reduce these burdens and democratize modern neuroscience research across institutions and career levels. Using community software and hardware infrastructure, the platform provides open-source data standardization, management, visualization, and processing and simplifies the data pipeline. brainlife.io automatically tracks the provenance history of thousands of data objects, supporting simplicity, efficiency, and transparency in neuroscience research. Here brainlife.io's technology and data services are described and evaluated for validity, reliability, reproducibility, replicability, and scientific utility. Using data from 4 modalities and 3,200 participants, we demonstrate that brainlife.io's services produce outputs that adhere to best practices in modern neuroscience research.

In astronomy or biological imaging, refractive index inhomogeneities of e.g. atmosphere or tissues induce optical aberrations which degrade the desired information hidden behind the medium. A standard approach consists in measuring these aberrations with a wavefront sensor (e.g Shack-Hartmann) located in the pupil plane, and compensating them either digitally or by adaptive optics with a wavefront shaper. However, in its usual implementation this strategy can only extract aberrations within a single isoplanatic patch, i.e. a region where the aberrations remain correlated. This limitation severely reduces the effective field-of-view in which the correction can be performed. Here, we propose a new wavefront sensing method capable of measuring, in a single shot, various pupil aberrations corresponding to multiple isoplanatic patches. The method, based on a thin diffuser (i.e a random phase mask), exploits the dissimilarity between different speckle regions to multiplex several wavefronts incoming from various incidence angles. We present proof-of-concept experiments carried out in wide-field fluorescence microscopy. A digital deconvolution procedure in each isoplanatic patch yields accurate aberration correction within an extended field-of-view. This approach is of interest for adaptive optics applications as well as diffractive optical tomography., Comment: 21 pages, 12 figures

The ability to automatically learn movements and behaviors of increasing complexity is a long-term goal in autonomous systems. Indeed, this is a very complex problem that involves understanding how knowledge is acquired and reused by humans as well as proposing mechanisms that allow artificial agents to reuse previous knowledge. Inspired by Jean Piaget's theory's first three sensorimotor substages, this work presents a cognitive agent based on CONAIM (Conscious Attention-Based Integrated Model) that can learn procedures incrementally. Throughout the paper, we show the cognitive functions required in each substage and how adding new functions helps address tasks previously unsolved by the agent. Experiments were conducted with a humanoid robot in a simulated environment modeled with the Cognitive Systems Toolkit (CST) performing an object tracking task. The system is modeled using a single procedural learning mechanism based on Reinforcement Learning. The increasing agent's cognitive complexity is managed by adding new terms to the reward function for each learning phase. Results show that this approach is capable of solving complex tasks incrementally., Comment: Preprint submitted to IEEE Transactions on Cognitive and Developmental Systems

Flexible graphite (FG) with 1 - 1.2 g/cm$^3$ density is employed as beam energy absorber material in the CERN's Large Hadron Collider (LHC) beam dumping system. However, the increase of energy deposited expected for new HL-LHC (High-Luminosity LHC) design demanded for an improvement in reliability and safety of beam dumping devices, and the need for a calibrated material model suitable for high-level FE simulations has been prioritized. This work sets the basic knowledge to develop a material model for FG suitable to this aim. A review of the FG properties available in literature is first given, followed by FIB-SEM (Focused Ion Beam - Scanning Electron Microscopy) microstructure investigation and monotonic and cyclic uniaxial compression tests. Similarities with other well-known groups of materials such as crushable foams, crumpled materials and compacted powders have been discussed. A simple 1D phenomenological model has been used to fit the experimental stress-strain curves and the accuracy of the result supports the assumptions that the graphite-like microstructure and the crumpled meso-structure play the major role under out-of-plane uniaxial compression., Comment: Pre-print template, 57 pages, 14 figures

Aberrations and multiple scattering in biological tissues critically distort light beams into highly complex speckle patterns. In this regard, digital optical phase conjugation (DOPC) is a promising technique enabling in-depth focusing. However, DOPC becomes challenging when using fluorescent guide-stars for four main reasons: The low photon budget available, the large spectral bandwidth of the fluorescent signal, the Stokes shift between the emission and the excitation wavelength, and the absence of reference beam preventing holographic measurement. Here, we demonstrate the possibility to focus a laser beam through multiple-scattering samples by measuring speckle fields in a single acquisition step with a reference-free, high-resolution wavefront sensor. By taking advantage of the large spectral bandwidth of forward multiply scattering samples, Digital Fluorescence Phase Conjugation (DFPC) is achieved to focus a laser beam at the excitation wavelength while measuring the broadband speckle field arising from a micron-sized fluorescent bead., Comment: 21 pages, 31 figures, 2 tables

Many processes in microfluidics and biology are driven or affected by viscosity. While several methods are able to measure this parameter globally (AFM, surface acoustic waves, DLS, ...), very few can provide high resolution viscosity images, particularly in living cells. Here, we propose to use sub-micrometer magnetic rods to perform high resolution viscosity imaging. An external magnetic field forces the oscillation of superparamagnetic iron oxide nanorods. Under linearly polarized illumination, the rotation of these highly anisotropic optical scatterers induces a blinking which is analyzed by heterodyne holography. The spectral analysis of the rotation dynamics yields a regime transition frequency f t from which the local viscosity is deduced. Holography provides 3D image reconstruction and 3D superlocalization of the nanorods, which allows super-resolved viscosity measurements. Relying on the fast Brownian rotation of nanorods instead of their slower translation therefore allows faster measurements and, crucially, smaller effective voxels for viscosity determination. Viscosity imaging is demonstrated with a 0.5 {\mu}m 3 3D resolution., Comment: 17 pages, 6 figures, Supplementary Information