Search

Composed Image Retrieval (CIR) provides an effective way to manage and access large-scale visual data. Construction of the CIR model utilizes triplets that consist of a reference image, modification text describing desired changes, and a target image that reflects these changes. For effectively training CIR models, extensive manual annotation to construct high-quality training datasets, which can be time-consuming and labor-intensive, is required. To deal with this problem, this paper proposes a novel triplet synthesis method by leveraging counterfactual image generation. By controlling visual feature modifications via counterfactual image generation, our approach automatically generates diverse training triplets without any manual intervention. This approach facilitates the creation of larger and more expressive datasets, leading to the improvement of CIR model's performance., Comment: 4 pages, 4 figures

Singular learning models with non-positive Fisher information matrices include neural networks, reduced-rank regression, Boltzmann machines, normal mixture models, and others. These models have been widely used in the development of learning machines. However, theoretical analysis is still in its early stages. In this paper, we examine learning coefficients, which indicate the general learning efficiency of deep linear learning models and three-layer neural network models with ReLU units. Finally, we extend the results to include the case of the Softmax function., Comment: 12 pages

As global attention on sustainable and ethical food systems grows, animal welfare-friendly products (AWFP) are increasingly recognized as essential to addressing consumer and producer concerns. However, traditional research often neglects the interdependencies between production, retail, and consumption stages within the supply chain. This study examined how cross-stage interactions among producers, consumers, and retail intermediaries can promote AWFP adoption. By establishing a short value chain from production to consumption, we conducted a two-month choice experiment in the operational restaurant, employing a mixed-method approach to quantitatively and qualitatively assess stakeholder responses. The results revealed that providing information about AWFP practices significantly influenced consumer behavior, increasing both product selection and perceived value. Retailers recognized the potential for economic benefits and strengthened customer loyalty, while producers identified new revenue opportunities by re-fattening delivered cow. These coordinated changes - defined as synchronized actions and mutual reinforcement across production, retail, and consumption - generated positive feedback loops that motivated stakeholders to adopt AWFP practices. This research underscores the potential of strategically designed short value chain to foster cross-stage coordination and highlights their role as practical entry points for promoting sustainable and ethical food systems on a larger scale., Comment: 26 pages, 6 figures, 10 tables

Device driver bugs are the leading cause of OS compromises, and their formal verification is therefore highly desirable. To the best of our knowledge, no realistic and performant driver has been verified for a non-trivial device. We propose Pancake, an imperative language for systems programming that features a well-defined and verification-friendly semantics. Leveraging the verified compiler backend of the CakeML functional language, we develop a compiler for Pancake that guarantees that the binary retains the semantics of the source code. Usng automatic translation of Pancake to the Viper SMT front-end, we verify a performant driver for an Ethernet NIC., Comment: 15 pages, 5 figures

We propose a novel continual self-supervised learning method (CSSL) considering medical domain knowledge in chest CT images. Our approach addresses the challenge of sequential learning by effectively capturing the relationship between previously learned knowledge and new information at different stages. By incorporating an enhanced DER into CSSL and maintaining both diversity and representativeness within the rehearsal buffer of DER, the risk of data interference during pretraining is reduced, enabling the model to learn more richer and robust feature representations. In addition, we incorporate a mixup strategy and feature distillation to further enhance the model's ability to learn meaningful representations. We validate our method using chest CT images obtained under two different imaging conditions, demonstrating superior performance compared to state-of-the-art methods., Comment: Accepted by ICASSP 2025

Dataset distillation is an effective technique for reducing the cost and complexity of model training while maintaining performance by compressing large datasets into smaller, more efficient versions. In this paper, we present a novel generative dataset distillation method that can improve the accuracy of aligning prediction logits. Our approach integrates self-knowledge distillation to achieve more precise distribution matching between the synthetic and original data, thereby capturing the overall structure and relationships within the data. To further improve the accuracy of alignment, we introduce a standardization step on the logits before performing distribution matching, ensuring consistency in the range of logits. Through extensive experiments, we demonstrate that our method outperforms existing state-of-the-art methods, resulting in superior distillation performance., Comment: Accepted by ICASSP 2025

We theoretically investigate the flow of information in an interacting two-skyrmion system confined in a box at finite temperature. Using numerical simulations based on the Thiele-Langevin equation, we demonstrate that the skyrmion motion cannot be fully described by the master equation, highlighting the system's simplicity with its nontrivial dynamics. Particularly, due to the chiral motion of skyrmion, we find asymmetric flow of information even in equilibrium, which is demonstrated by the violation of the detailed balance condition. We analyze this novel system using information-theoretical quantities including Shannon entropy, mutual information, and transfer entropy. By the analyses, the physical significance of transfer entropy as a measure of flow of information, which has been overlooked in previous studies, is elucidated. Notably, the peak position of the transfer entropy, as a function of time delay, is found to be independent of the interaction range between the two skyrmions yet dependent on the box size. This peak corresponds to the characteristic time required for changing the skyrmion state: the box size divided by the average velocity of skyrmions. We can understand that the information transmission time consists of the time to obtain mutual information and the time to write the information. Since the unusual asymmetric circulation of information is revealed in this two-skyrmion system, it can be a unique platform for future applications to natural computing using flow of information, including more efficient machine learning algorithm., Comment: 13 pages, 8 figures

Knowledge Graphs (KGs) represent relationships between entities in a graph structure and have been widely studied as promising tools for realizing recommendations that consider the accurate content information of items. However, traditional KG-based recommendation methods face fundamental challenges: insufficient consideration of temporal information and poor performance in cold-start scenarios. On the other hand, Large Language Models (LLMs) can be considered databases with a wealth of knowledge learned from the web data, and they have recently gained attention due to their potential application as recommendation systems. Although approaches that treat LLMs as recommendation systems can leverage LLMs' high recommendation literacy, their input token limitations make it impractical to consider the entire recommendation domain dataset and result in scalability issues. To address these challenges, we propose a LLM's Intuition-aware Knowledge graph Reasoning model (LIKR). Our main idea is to treat LLMs as reasoners that output intuitive exploration strategies for KGs. To integrate the knowledge of LLMs and KGs, we trained a recommendation agent through reinforcement learning using a reward function that integrates different recommendation strategies, including LLM's intuition and KG embeddings. By incorporating temporal awareness through prompt engineering and generating textual representations of user preferences from limited interactions, LIKR can improve recommendation performance in cold-start scenarios. Furthermore, LIKR can avoid scalability issues by using KGs to represent recommendation domain datasets and limiting the LLM's output to KG exploration strategies. Experiments on real-world datasets demonstrate that our model outperforms state-of-the-art recommendation methods in cold-start sequential recommendation scenarios., Comment: Accepted to the 47th European Conference on Information Retrieval (ECIR2025)

In few-shot action recognition (FSAR), long sub-sequences of video naturally express entire actions more effectively. However, the high computational complexity of mainstream Transformer-based methods limits their application. Recent Mamba demonstrates efficiency in modeling long sequences, but directly applying Mamba to FSAR overlooks the importance of local feature modeling and alignment. Moreover, long sub-sequences within the same class accumulate intra-class variance, which adversely impacts FSAR performance. To solve these challenges, we propose a Matryoshka MAmba and CoNtrasTive LeArning framework (Manta). Firstly, the Matryoshka Mamba introduces multiple Inner Modules to enhance local feature representation, rather than directly modeling global features. An Outer Module captures dependencies of timeline between these local features for implicit temporal alignment. Secondly, a hybrid contrastive learning paradigm, combining both supervised and unsupervised methods, is designed to mitigate the negative effects of intra-class variance accumulation. The Matryoshka Mamba and the hybrid contrastive learning paradigm operate in two parallel branches within Manta, enhancing Mamba for FSAR of long sub-sequence. Manta achieves new state-of-the-art performance on prominent benchmarks, including SSv2, Kinetics, UCF101, and HMDB51. Extensive empirical studies prove that Manta significantly improves FSAR of long sub-sequence from multiple perspectives., Comment: Accepted by AAAI 2025

GPU is the dominant accelerator device due to its high performance and energy efficiency. Directive-based GPU offloading using OpenACC or OpenMP target is a convenient way to port existing codes originally developed for multicore CPUs. Although OpenACC and OpenMP target provide similar features, both methods have pros and cons. OpenACC has better functions and an abundance of documents, but it is virtually for NVIDIA GPUs. OpenMP target supports NVIDIA/AMD/Intel GPUs but has fewer functions than OpenACC. Here, we have developed a header-only library, Solomon (Simple Off-LOading Macros Orchestrating multiple Notations), to unify the interface for GPU offloading with the support of both OpenACC and OpenMP target. Solomon provides three types of notations to reduce users' implementation and learning costs: intuitive notation for beginners and OpenACC/OpenMP-like notations for experienced developers. This manuscript denotes Solomon's implementation and usage and demonstrates the GPU-offloading in $N$-body simulation and the three-dimensional diffusion equation. The library and sample codes are provided as open-source software and publicly and freely available at \url{https://github.com/ymiki-repo/solomon}., Comment: 24 pages, 2 figures, 21 tables, accepted for publication in IEEE Access. The library and sample codes are available at https://github.com/ymiki-repo/solomon

In photonic quantum information processing, quantum operations using nonlinear photon-photon interactions are vital for implementing two-qubit gates and enabling faithful entanglement swapping. However, due to the weak interaction between single photons, the all-photonic realization of such quantum operations has remained out of reach so far. Herein, we demonstrate a first entanglement swapping using sum-frequency generation (SFG) between single photons in a $\chi^{(2)}$-nonlinear optical waveguide. We show that a highly efficient, stable SFG-based Bell-state analyzer and an ultralow-dark-count superconducting single-photon detector satisfy the high signal-to-noise ratio requirement for the swapping protocol.Furthermore, the system clock is enhanced by utilizing ultrafast telecom entangled photon pair sources that operate in the GHz range. Our results confirm a lower bound 0.770(76) for the swapped state's fidelity, surpassing the classical limit of 0.5 successfully. Our findings highlight the strong potential of broadband all-single-photonic nonlinear interactions for further sophistication in long-distance quantum communication and photonic quantum computation., Comment: 17 pages, 4 figures

Numerous wearable robots have been developed to meet the demands of physical assistance and entertainment. These wearable robots range from body-enhancing types that assist human arms and legs to body-extending types that have extra arms. This study focuses specifically on wearable robots of the latter category, aimed at bodily extension. However, they have not yet achieved the level of powerfulness and reachability equivalent to that of human limbs, limiting their application to entertainment and manipulation tasks involving lightweight objects. Therefore, in this study, we develop an body-extending wearable robot, Vlimb, which has enough powerfulness to lift a human and can perform manipulation. Leveraging the advantages of tendon-driven mechanisms, Vlimb incorporates a wire routing mechanism capable of accommodating both delicate manipulations and robust lifting tasks. Moreover, by introducing a passive ring structure to overcome the limited reachability inherent in tendon-driven mechanisms, Vlimb achieves both the powerfulness and reachability comparable to that of humans. This paper outlines the design methodology of Vlimb, conducts preliminary manipulation and lifting tasks, and verifies its effectiveness.

Quantifying interaction strengths between state variables in dynamical systems is essential for understanding ecological networks. Within the empirical dynamic modeling approach, multivariate S-map infers the interaction Jacobian from time series data without assuming specific dynamical models. This approach enables the non-parametric statistical inference of interspecific interactions through state space reconstruction. However, deviations in the biological interpretation and numerical implementation of the interaction Jacobian from its mathematical definition pose challenges. We mathematically reintroduce the interaction Jacobian using differential quotients, uncovering two problems: (1) the mismatch between the interaction Jacobian and its biological meaning complicates comparisons between interspecific and intraspecific interactions; (2) the interaction Jacobian is not fully implemented in the parametric Jacobian numerically derived from given parametric models, especially using ordinary differential equations. As a result, model-based evaluations of S-map methods become inappropriate. To address these problems, (1) we propose adjusting the diagonal elements of the interaction Jacobian by subtracting 1 to resolve the comparability problem between inter- and intraspecific interaction strengths. Simulations of population dynamics showed that this adjustment prevents overestimation of intraspecific interaction strengths. (2) We introduce an alternative parametric Jacobian and then cumulative interaction strength (CIS), providing a more rigorous benchmark for evaluating S-map methods. Furthermore, we demonstrated that the numerical gap between CIS and the existing parametric Jacobian is substantial in realistic scenarios, suggesting CIS as preferred benchmark. These solutions offer a clearer framework for developing non-parametric approaches in ecological time series analysis., Comment: 39 pages, 5 figures and 2 tables

In histological pathology, frozen sections are often used for rapid diagnosis during surgeries, as they can be produced within minutes. However, they suffer from artifacts and often lack crucial diagnostic details, particularly within the cell nuclei region. Permanent sections, on the other hand, contain more diagnostic detail but require a time-intensive preparation process. Here, we present a generative deep learning approach to enhance frozen section images by leveraging guidance from permanent sections. Our method places a strong emphasis on the nuclei region, which contains critical information in both frozen and permanent sections. Importantly, our approach avoids generating artificial data in blank regions, ensuring that the network only enhances existing features without introducing potentially unreliable information. We achieve this through a segmented attention network, incorporating nuclei-segmented images during training and adding an additional loss function to refine the nuclei details in the generated permanent images. We validated our method across various tissues, including kidney, breast, and colon. This approach significantly improves histological efficiency and diagnostic accuracy, enhancing frozen section images within seconds, and seamlessly integrating into existing laboratory workflows.

One of the important advantages of musculoskeletal humanoids is that the muscle arrangement can be easily changed and the number of muscles can be increased according to the situation. In this study, we describe an overall system of muscle addition for musculoskeletal humanoids and the adaptive body schema learning while taking into account the additional muscles. For hardware, we describe a modular body design that can be fitted with additional muscles, and for software, we describe a method that can learn the changes in body schema associated with additional muscles from a small amount of motion data. We apply our method to a simple 1-DOF tendon-driven robot simulation and the arm of the musculoskeletal humanoid Musashi, and show the effectiveness of muscle tension relaxation by adding muscles for a high-load task., Comment: Accepted at IEEE Robotics and Automation Letters

Short video platforms, such as YouTube, Instagram, or TikTok, are used by billions of users globally. These platforms expose users to harmful content, ranging from clickbait or physical harms to misinformation or online hate. Yet, detecting harmful videos remains challenging due to an inconsistent understanding of what constitutes harm and limited resources and mental tolls involved in human annotation. As such, this study advances measures and methods to detect harm in video content. First, we develop a comprehensive taxonomy for online harm on video platforms, categorizing it into six categories: Information, Hate and harassment, Addictive, Clickbait, Sexual, and Physical harms. Next, we establish multimodal large language models as reliable annotators of harmful videos. We analyze 19,422 YouTube videos using 14 image frames, 1 thumbnail, and text metadata, comparing the accuracy of crowdworkers (Mturk) and GPT-4-Turbo with domain expert annotations serving as the gold standard. Our results demonstrate that GPT-4-Turbo outperforms crowdworkers in both binary classification (harmful vs. harmless) and multi-label harm categorization tasks. Methodologically, this study extends the application of LLMs to multi-label and multi-modal contexts beyond text annotation and binary classification. Practically, our study contributes to online harm mitigation by guiding the definitions and identification of harmful content on video platforms.

For robots to become more versatile and expand their areas of application, their bodies need to be suitable for contact with the environment. When the human body comes into contact with the environment, it is possible for it to continue to move even if the positional relationship between muscles or the shape of the muscles changes. We have already focused on the effect of geometric deformation of muscles and proposed a drive system called wire-wound Muscle-Tendon Complex (ww-MTC), an extension of the wire drive system. Our previous study using a robot with a two-dimensional configuration demonstrated several advantages: reduced wire loosening, interference, and wear; improved robustness during environmental contact; and a muscular appearance. However, this design had some problems, such as excessive muscle expansion that hindered inter-muscle movement, and confinement to planar motion. In this study, we develop the ww-MTC into a three-dimensional shape. We present a fundamental construction method for a muscle exterior that expands gently and can be contacted over its entire surface. We also apply the three-dimensional ww-MTC to a 2-axis 3-muscle robot, and confirm that the robot can continue to move while adapting to its environment., Comment: Accepted at Humanoids2024, website - https://sites.google.com/view/yoshimoto-ribayashi/projects, YouTube - https://youtu.be/EDeAqg7aAb4

Chiral and axial materials offer platforms for intriguing phenomena, such as cross-correlated responses and chirality-induced spin selectivity. However, quantifying the properties of such materials has generally been considered challenging. Here, we demonstrate that the spatial distribution of the electron chirality, represented by $\Psi^\dagger \gamma^5 \Psi$ with the four-component Dirac field $\Psi$, characterizes the chirality and axiality of materials. Furthermore, we reveal that spin-derived electric polarization can serve as an effective indicator of material polarity. We present quantitative evaluations of electron chirality distribution and spin-derived electric polarization based on first-principles calculations. Additionally, we propose that electron chirality can be directly observed via circular dichroism in photoemission spectroscopy, which measures the difference between right- and left-handed circularly polarized light. Electron chirality and spin-derived electric polarization provide a new framework for quantifying chirality, axiality, and polarity in asymmetric materials, paving the way for the exploration of novel functional materials., Comment: 9 pages, 7 figures

Dimensionality reduction (DR) offers a useful representation of complex high-dimensional data. Recent DR methods focus on hyperbolic geometry to derive a faithful low-dimensional representation of hierarchical data. However, existing methods are based on neighbor embedding, frequently ruining the continual relation of the hierarchies. This paper presents hyperboloid Gaussian process (GP) latent variable models (hGP-LVMs) to embed high-dimensional hierarchical data with implicit continuity via nonparametric estimation. We adopt generative modeling using the GP, which brings effective hierarchical embedding and executes ill-posed hyperparameter tuning. This paper presents three variants that employ original point, sparse point, and Bayesian estimations. We establish their learning algorithms by incorporating the Riemannian optimization and active approximation scheme of GP-LVM. For Bayesian inference, we further introduce the reparameterization trick to realize Bayesian latent variable learning. In the last part of this paper, we apply hGP-LVMs to several datasets and show their ability to represent high-dimensional hierarchies in low-dimensional spaces.

Conventional medical artificial intelligence (AI) models face barriers in clinical application and ethical issues owing to their inability to handle the privacy-sensitive characteristics of medical data. We present a novel personalized federated learning (pFL) method for medical visual question answering (VQA) models, addressing privacy reliability challenges in the medical domain. Our method introduces learnable prompts into a Transformer architecture to efficiently train it on diverse medical datasets without massive computational costs. Then we introduce a reliable client VQA model that incorporates Dempster-Shafer evidence theory to quantify uncertainty in predictions, enhancing the model's reliability. Furthermore, we propose a novel inter-client communication mechanism that uses maximum likelihood estimation to balance accuracy and uncertainty, fostering efficient integration of insights across clients.

As the global food system faces increasing challenges from sustainability, climate change, and food security issues, alternative food networks like Community-Supported Agriculture (CSA) play an essential role in fostering stronger connections between consumers and producers. However, understanding consumer engagement with CSA is fragmented, particularly in Japan where CSA participation is still emerging. This study aims to identify potential CSA participants in Japan and validate existing theories on CSA participation through a quantitative analysis of 2,484 Japanese consumers. Using choice experiments, Latent Class Analysis, and Partial Least Squares Structural Equation Modeling, we identified five distinct consumer segments. The "Sustainable Food Seekers" group showed the highest positive utility for CSA, driven primarily by "Food Education and Learning Opportunities" and "Contribution to Environmental and Social Issues." These factors were consistently significant across all segments, suggesting that many Japanese consumers value CSA for its educational and environmental benefits. In contrast, factors related to "Variety of Ingredients" were less influential in determining participation intentions. The findings suggest that promoting CSA in Japan may be most effective by emphasizing its role in environmental and social impact, rather than focusing solely on product attributes like organic certification, which is readily available in supermarkets. This reflects a key distinction between CSA adoption in Japan and in other cultural contexts, where access to organic produce is a primary driver. For "Sustainable Food Seekers," CSA offers a way to contribute to broader societal goals rather than just securing organic products., Comment: 29 pages, 5 figures

Muscles of the human body are composed of tiny actuators made up of myosin and actin filaments. They can exert force in various shapes such as curved or flat, under contact forces and deformations from the environment. On the other hand, muscles in musculoskeletal robots so far have faced challenges in generating force in such shapes and environments. To address this issue, we propose Patterned Structure Muscle (PSM), artificial muscles for musculoskeletal robots. PSM utilizes patterned structures with anisotropic characteristics, wire-driven mechanisms, and is made of flexible material Thermoplastic Polyurethane (TPU) using FDM 3D printing. This method enables the creation of various shapes of muscles, such as simple 1 degree-of-freedom (DOF) muscles, Multi-DOF wide area muscles, joint-covering muscles, and branched muscles. We created an upper arm structure using these muscles to demonstrate wide range of motion, lifting heavy objects, and movements through environmental contact. These experiments show that the proposed PSM is capable of operating in various shapes and environments, and is suitable for the muscles of musculoskeletal robots., Comment: accepted at IROS2024

The complex ways in which humans utilize their bodies in sports and martial arts are remarkable, and human motion analysis is one of the most effective tools for robot body design and control. On the other hand, motion analysis is not easy, and it is difficult to measure complex body motions in detail due to the influence of numerous muscles and soft tissues, mainly ligaments. In response, various musculoskeletal simulators have been developed and applied to motion analysis and robotics. However, none of them reproduce the ligaments but only the muscles, nor do they focus on the shoulder complex, including the clavicle and scapula, which is one of the most complex parts of the body. Therefore, in this study, a detailed simulation model of the shoulder complex including ligaments is constructed. The model will mimic not only the skeletal structure and muscle arrangement but also the ligament arrangement and maximum muscle strength. Through model predictive control based on the constructed simulation, we confirmed that the ligaments contribute to joint stabilization in the first movement and that the proper distribution of maximum muscle force contributes to the equalization of the load on each muscle, demonstrating the effectiveness of this simulation., Comment: accepted at IROS2024, websites - https://sahara-yuta.github.io/projects/shoulder-complex-simulation

A search for proton decay into $e^+/\mu^+$ and a $\eta$ meson has been performed using data from a 0.373 Mton$\cdot$year exposure (6050.3 live days) of Super-Kamiokande. Compared to previous searches this work introduces an improved model of the intranuclear $\eta$ interaction cross section, resulting in a factor of two reduction in uncertainties from this source and $\sim$10\% increase in signal efficiency. No significant data excess was found above the expected number of atmospheric neutrino background events resulting in no indication of proton decay into either mode. Lower limits on the proton partial lifetime of $1.4\times\mathrm{10^{34}~years}$ for $p\rightarrow e^+\eta$ and $7.3\times\mathrm{10^{33}~years}$ for $p\rightarrow \mu^+\eta$ at the 90$\%$ C.L. were set. These limits are around 1.5 times longer than our previous study and are the most stringent to date.

Dynamic manipulation of flexible objects such as fabric, which is difficult to modelize, is one of the major challenges in robotics. With the development of deep learning, we are beginning to see results in simulations and in some actual robots, but there are still many problems that have not yet been tackled. Humans can move their arms at high speed using their flexible bodies skillfully, and even when the material to be manipulated changes, they can manipulate the material after moving it several times and understanding its characteristics. Therefore, in this research, we focus on the following two points: (1) body control using a variable stiffness mechanism for more dynamic manipulation, and (2) response to changes in the material of the manipulated object using parametric bias. By incorporating these two approaches into a deep predictive model, we show through simulation and actual robot experiments that Musashi-W, a musculoskeletal humanoid with variable stiffness mechanism, can dynamically manipulate cloth while detecting changes in the physical properties of the manipulated object., Comment: Accepted at Frontiers in Neurorobotics

Intermittent switchings between weakly chaotic (laminar) and strongly chaotic (bursty) states are often observed in systems with high-dimensional chaotic attractors, such as fluid turbulence. They differ from the intermittency of a low-dimensional system accompanied by the stability change of a fixed point or a periodic orbit in that the intermittency of a high-dimensional system tends to appear in a wide range of parameters. This paper considers a case where the skeleton of a laminar state $L$ exists as a proper chaotic subset $S$ of a chaotic attractor $X$, that is, $S\ \subsetneq\ X$. We characterize such a laminar state $L$ by a chaotic saddle $S$, which is densely filled with periodic orbits of different numbers of unstable directions. This study demonstrates the presence of chaotic saddles underlying intermittency in fluid turbulence and phase synchronization. Furthermore, we confirm that chaotic saddles persist for a wide range of parameters. Also, a kind of phase synchronization turns out to occur in the turbulent model., Comment: 7 pages, 7 figures

Dense Wavelength Division Multiplexing (DWDM) is a key technology for realizing high-capacity and flexible quantum communication networks. In addition, to realize the emerging quantum internet, quantum frequency conversion is also essential for bridging different quantum systems over optical fiber networks. In this work, we demonstrate a channel-selective quantum frequency conversion (CS-QFC), which allows active selection of the frequency of the converted photon from multiple DWDM channels. The 2.5 THz bandwidth of our CS-QFC system shows the ability to establish a 100-ch DWDM dynamic link from a single quantum system. It promises to increase the diversity of the quantum network., Comment: 6 pages, 4 figures

Earth's geodynamo has operated for over 3.5 billion years. The magnetic field is currently powered by thermocompositional convection in the outer core, which involves the release of light elements and latent heat as the inner core solidifies. However, since the inner core nucleated no more than 1.5 billion years ago, the early dynamo could not rely on these buoyancy sources. Given recent estimates of the thermal conductivity of the outer core, an alternative mechanism may be required to sustain the geodynamo prior to nucleation of the inner core. One possibility is a silicate dynamo operating in a long-lived basal magma ocean. Here, we investigate the structural, thermal, buoyancy, and magnetic evolution of an Earth-like terrestrial planet. Using modern equations of state and melting curves, we include a time-dependent parameterization of the compositional evolution of an iron-rich basal magma ocean. We combine an internal structure integration of the planet with energy budgets in a coupled core, basal magma ocean, and mantle system. We determine the thermocompositional convective stability of the core and the basal magma ocean, and assess their respective dynamo activity using entropy budgets and magnetic Reynolds numbers. Our conservative nominal model predicts a transient basal magma ocean dynamo followed by a core dynamo after 1 billion years. The model is sensitive to several parameters, including the initial temperature of the core-mantle boundary, the parameterization of mantle convection, the composition of the basal magma ocean, the radiogenic content of the planet, as well as convective velocity and magnetic scaling laws. We use the nominal model to constrain the range of basal magma ocean electrical conductivity and core thermal conductivity that sustain a dynamo.

Quantum frequency conversion (QFC) which converts the frequencies of photons while preserving the quantum state is an essential technology for realizing the quantum internet and quantum interconnect. For the QFC based on the frequency downconversion from visible to the telecom wavelengths around 1500 nm, it is widely known that noise photons produced by the strong pump light used for QFC contaminate the frequency-converted photon, which degrades the quality of the quantum property of the photon after QFC. In conventional QFC experiments, noise photons are removed using external narrowband frequency filter systems. In contrast, in this study, we implement a compact QFC device integrating the cavity structure only for the converted mode. While the cavity structure can enhance not only the desired QFC efficiency but also the noise photon generation rate, we show that the cavity-enhanced QFC followed by a relatively wide bandpass filter achieves the signal-to-noise ratio comparable to the QFCs with external narrowband filters. We experimentally demonstrate the cavity-enhanced QFC using a single photon at 780 nm to 1540 nm, in which the non-classical photon statistics is clearly observed after QFC., Comment: 10 pages, 12 figures

Using doors is a longstanding challenge in robotics and is of significant practical interest in giving robots greater access to human-centric spaces. The task is challenging due to the need for online adaptation to varying door properties and precise control in manipulating the door panel and navigating through the confined doorway. To address this, we propose a learning-based controller for a legged manipulator to open and traverse through doors. The controller is trained using a teacher-student approach in simulation to learn robust task behaviors as well as estimate crucial door properties during the interaction. Unlike previous works, our approach is a single control policy that can handle both push and pull doors through learned behaviour which infers the opening direction during deployment without prior knowledge. The policy was deployed on the ANYmal legged robot with an arm and achieved a success rate of 95.0% in repeated trials conducted in an experimental setting. Additional experiments validate the policy's effectiveness and robustness to various doors and disturbances. A video overview of the method and experiments can be found at youtu.be/tQDZXN_k5NU.

We propose a new strategy called think twice before recognizing to improve fine-grained traffic sign recognition (TSR). Fine-grained TSR in the wild is difficult due to the complex road conditions, and existing approaches particularly struggle with cross-country TSR when data is lacking. Our strategy achieves effective fine-grained TSR by stimulating the multiple-thinking capability of large multimodal models (LMM). We introduce context, characteristic, and differential descriptions to design multiple thinking processes for the LMM. The context descriptions with center coordinate prompt optimization help the LMM to locate the target traffic sign in the original road images containing multiple traffic signs and filter irrelevant answers through the proposed prior traffic sign hypothesis. The characteristic description is based on few-shot in-context learning of template traffic signs, which decreases the cross-domain difference and enhances the fine-grained recognition capability of the LMM. The differential descriptions of similar traffic signs optimize the multimodal thinking capability of the LMM. The proposed method is independent of training data and requires only simple and uniform instructions. We conducted extensive experiments on three benchmark datasets and two real-world datasets from different countries, and the proposed method achieves state-of-the-art TSR results on all five datasets.

We propose a novel symbolic music representation and Generative Adversarial Network (GAN) framework specially designed for symbolic multitrack music generation. The main theme of symbolic music generation primarily encompasses the preprocessing of music data and the implementation of a deep learning framework. Current techniques dedicated to symbolic music generation generally encounter two significant challenges: training data's lack of information about chords and scales and the requirement of specially designed model architecture adapted to the unique format of symbolic music representation. In this paper, we solve the above problems by introducing new symbolic music representation with MusicLang chord analysis model. We propose our MMT-BERT architecture adapting to the representation. To build a robust multitrack music generator, we fine-tune a pre-trained MusicBERT model to serve as the discriminator, and incorporate relativistic standard loss. This approach, supported by the in-depth understanding of symbolic music encoded within MusicBERT, fortifies the consonance and humanity of music generated by our method. Experimental results demonstrate the effectiveness of our approach which strictly follows the state-of-the-art methods., Comment: Accepted to the 25th International Society for Music Information Retrieval Conference (ISMIR 2024)

We give explicit formulas of Whittaker functions on GL(4,R) for all irreducible generic representations. As an application, we determine test vectors which attain the associated L-factors for Bump-Friedberg integrals on GL(4,R).

Identifying the physical properties of the surrounding environment is essential for robotic locomotion and navigation to deal with non-geometric hazards, such as slippery and deformable terrains. It would be of great benefit for robots to anticipate these extreme physical properties before contact; however, estimating environmental physical parameters from vision is still an open challenge. Animals can achieve this by using their prior experience and knowledge of what they have seen and how it felt. In this work, we propose a cross-modal self-supervised learning framework for vision-based environmental physical parameter estimation, which paves the way for future physical-property-aware locomotion and navigation. We bridge the gap between existing policies trained in simulation and identification of physical terrain parameters from vision. We propose to train a physical decoder in simulation to predict friction and stiffness from multi-modal input. The trained network allows the labeling of real-world images with physical parameters in a self-supervised manner to further train a visual network during deployment, which can densely predict the friction and stiffness from image data. We validate our physical decoder in simulation and the real world using a quadruped ANYmal robot, outperforming an existing baseline method. We show that our visual network can predict the physical properties in indoor and outdoor experiments while allowing fast adaptation to new environments.

Inspired by recent developments in electron chirality, we reconsider some microscopic physical quantities that have been overlooked or have received little attention in condensed matter physics, based on the non-relativistic limit of the Dirac bilinears in relativistic quantum theory. We identify the expression of physical quantities defined by the four-component Dirac field in terms of the two-component Schr\"odinger field, which is usually used in condensed matter physics, and clarify its conjugate electromagnetic field. This consideration bridges the fields of condensed matter physics, quantum chemistry, and particle physics, and paves the way to electromagnetic control of matter. Our findings provide a means of {\it ab initio} quantification of material characters such as chirality and axiality that are unique to low-symmetry materials, and stimulate the systematic search for useful, new functionalities., Comment: 22 pages, 2 figures

Here we critically examine the geophysical and geochemical properties of the Moon in order to identify the extent to which dynamical scenarios satisfy these observations. New joint inversions of existing lunar geophysical data (mean mass, moment of inertia, and tidal response) assuming a laterally- and vertically homogeneous lunar mantle show that, in all cases, a core with a radius of 300$\pm$20 km ($\sim$0.8 to 1.5 % the mass of the Moon) is required. However, an Earth-like Mg# (0.89) in the lunar mantle results in core densities (7800$\pm$100 kg/m$^3$) consistent with that of Fe-Ni alloy, whereas FeO-rich compositions (Mg# = 0.80--0.84) require lower densities (6100$\pm$800 kg/m$^3$). Geochemically, we use new data on mare basalts to reassess the bulk composition of the Moon for 70 elements, and show that the lunar core likely formed near 5 GPa, 2100 K and $\sim$1 log unit below the iron-w\"ustite buffer. Moreover, the Moon is depleted relative to the Earth's mantle in elements with volatilities higher than that of Li, with this volatile loss likely having occurred at low temperatures (1400$\pm$100 K), consistent with mass-dependent stable isotope fractionation of moderately volatile elements (e.g., Zn, K, Rb). The identical nucleosynthetic (O, Cr, Ti) and radiogenic (W) isotope compositions of the lunar and terrestrial mantles, strongly suggest the two bodies were made from the same material, rather than from an Earth-like impactor. Rb-Sr in FANs and Lu-Hf and Pb-Pb zircon ages point Moon formation close to $\sim$4500 Ma. Taken together, there is no unambiguous geochemical or isotopic evidence for the role of an impactor in the formation of the Moon, implying perfect equilibration between the proto-Earth and Moon-forming material or alternative scenarios for its genesis., Comment: 62 pages, 23 figures, 5 tables. Treatise of Geochemistry, vol. 3, in press

Positive muon spin rotation and relaxation spectroscopy is a well established experimental technique for studying materials. It provides a local probe that generally complements scattering techniques in the study of magnetic systems and represents a valuable alternative for materials that display strong incoherent scattering or neutron absorption. Computational methods can effectively quantify the microscopic interactions underlying the experimentally observed signal, thus substantially boosting the predictive power of this technique. Here, we present an efficient set of algorithms and workflows devoted to the automation of this task. In particular, we adopt the so-called DFT+{\mu} procedure, where the system is characterised in the density functional theory (DFT) framework with the muon modeled as a hydrogen impurity. We devise an automated strategy to obtain candidate muon stopping sites, their dipolar interaction with the nuclei, and hyperfine interactions with the electronic ground state. We validate the implementation on well-studied compounds, showing the effectiveness of our protocol in terms of accuracy and simplicity of use

This paper presents our method for the generative track of The First Dataset Distillation Challenge at ECCV 2024. Since the diffusion model has become the mainstay of generative models because of its high-quality generative effects, we focus on distillation methods based on the diffusion model. Considering that the track can only generate a fixed number of images in 10 minutes using a generative model for CIFAR-100 and Tiny-ImageNet datasets, we need to use a generative model that can generate images at high speed. In this study, we proposed a novel generative dataset distillation method based on Stable Diffusion. Specifically, we use the SDXL-Turbo model which can generate images at high speed and quality. Compared to other diffusion models that can only generate images per class (IPC) = 1, our method can achieve an IPC = 10 for Tiny-ImageNet and an IPC = 20 for CIFAR-100, respectively. Additionally, to generate high-quality distilled datasets for CIFAR-100 and Tiny-ImageNet, we use the class information as text prompts and post data augmentation for the SDXL-Turbo model. Experimental results show the effectiveness of the proposed method, and we achieved third place in the generative track of the ECCV 2024 DD Challenge. Codes are available at https://github.com/Guang000/BANKO., Comment: The Third Place Winner in Generative Track of the ECCV 2024 DD Challenge

Objectives: To evaluate the zero-shot performance of Segment Anything Model 2 (SAM 2) in 3D segmentation of abdominal organs in CT scans, and to investigate the effects of prompt settings on segmentation results. Materials and Methods: In this retrospective study, we used a subset of the TotalSegmentator CT dataset from eight institutions to assess SAM 2's ability to segment eight abdominal organs. Segmentation was initiated from three different z-coordinate levels (caudal, mid, and cranial levels) of each organ. Performance was measured using the Dice similarity coefficient (DSC). We also analyzed the impact of "negative prompts," which explicitly exclude certain regions from the segmentation process, on accuracy. Results: 123 patients (mean age, 60.7 \pm 15.5 years; 63 men, 60 women) were evaluated. As a zero-shot approach, larger organs with clear boundaries demonstrated high segmentation performance, with mean DSCs as follows: liver 0.821 \pm 0.192, right kidney 0.862 \pm 0.212, left kidney 0.870 \pm 0.154, and spleen 0.891 \pm 0.131. Smaller organs showed lower performance: gallbladder 0.531 \pm 0.291, pancreas 0.361 \pm 0.197, and adrenal glands, right 0.203 \pm 0.222, left 0.308 \pm 0.234. The initial slice for segmentation and the use of negative prompts significantly influenced the results. By removing negative prompts from the input, the DSCs significantly decreased for six organs. Conclusion: SAM 2 demonstrated promising zero-shot performance in segmenting certain abdominal organs in CT scans, particularly larger organs. Performance was significantly influenced by input negative prompts and initial slice selection, highlighting the importance of optimizing these factors., Comment: 20 pages, 7 figures (including 2 supplemental figure), 4 tables

The neutron-rich unbound fluorine isotope $^{30}$F$_{21}$ has been observed for the first time by measuring its neutron decay at the SAMURAI spectrometer (RIBF, RIKEN) in the quasi-free proton knockout reaction of $^{31}$Ne nuclei at 235 MeV/nucleon. The mass and thus one-neutron-separation energy of $^{30}$F has been determined to be $S_n = -472\pm 58 \mathrm{(stat.)} \pm 33 \mathrm{(sys.)}$ keV from the measurement of its invariant-mass spectrum. The absence of a sharp drop in $S_n$($^{30}$F) shows that the ``magic'' $N=20$ shell gap is not restored close to $^{28}$O, which is in agreement with our shell-model calculations that predict a near degeneracy between the neutron $d$ and $fp$ orbitals, with the $1p_{3/2}$ and $1p_{1/2}$ orbitals becoming more bound than the $0f_{7/2}$ one. This degeneracy and reordering of orbitals has two potential consequences: $^{28}$O behaves like a strongly superfluid nucleus with neutron pairs scattering across shells, and both $^{29,31}$F appear to be good two-neutron halo-nucleus candidates., Comment: 9 pages, 2 figures, accepted for publication in Physical Review Letters

Recent multimodal large language models (MLLM) such as GPT-4o and GPT-4v have shown great potential in autonomous driving. In this paper, we propose a cross-domain few-shot in-context learning method based on the MLLM for enhancing traffic sign recognition (TSR). We first construct a traffic sign detection network based on Vision Transformer Adapter and an extraction module to extract traffic signs from the original road images. To reduce the dependence on training data and improve the performance stability of cross-country TSR, we introduce a cross-domain few-shot in-context learning method based on the MLLM. To enhance MLLM's fine-grained recognition ability of traffic signs, the proposed method generates corresponding description texts using template traffic signs. These description texts contain key information about the shape, color, and composition of traffic signs, which can stimulate the ability of MLLM to perceive fine-grained traffic sign categories. By using the description texts, our method reduces the cross-domain differences between template and real traffic signs. Our approach requires only simple and uniform textual indications, without the need for large-scale traffic sign images and labels. We perform comprehensive evaluations on the German traffic sign recognition benchmark dataset, the Belgium traffic sign dataset, and two real-world datasets taken from Japan. The experimental results show that our method significantly enhances the TSR performance.

The calcium isotopes are an ideal system to investigate the evolution of shell structure and magic numbers. Although the properties of surface nucleons in calcium have been well studied, probing the structure of deeply bound nucleons remains a challenge. Here, we report on the first measurement of unbound states in $^{53}$Ca and $^{55}$Ca, populated from \ts{54,56}Ca($p,pn$) reactions at a beam energy of around 216 MeV/nucleon at the RIKEN Radioactive Isotopes Beam Factory. The resonance properties, partial cross sections, and momentum distributions of these unbound states were analyzed. Orbital angular momentum $l$ assignments were extracted from momentum distributions based on calculations using the distorted wave impulse approximation (DWIA) reaction model. The resonances at excitation energies of 5516(41)\,keV in $^{53}$Ca and 6000(250)\,keV in $^{55}$Ca indicate a significant $l$\, =\,3 component, providing the first experimental evidence for the $\nu 0f_{7/2}$ single-particle strength of unbound hole states in the neutron-rich Ca isotopes. The observed excitation energies and cross-sections point towards extremely localized and well separated strength distributions, with some fragmentation for the $\nu 0f_{7/2}$ orbital in $^{55}$Ca. These results are in good agreement with predictions from shell-model calculations using the effective GXPF1Bs interaction and \textit{ab initio} calculations and diverge markedly from the experimental distributions in the nickel isotones at $Z=28$., Comment: 13 pages, 7 figures

This paper proposes a novel zero-shot composed image retrieval (CIR) method considering the query-target relationship by masked image-text pairs. The objective of CIR is to retrieve the target image using a query image and a query text. Existing methods use a textual inversion network to convert the query image into a pseudo word to compose the image and text and use a pre-trained visual-language model to realize the retrieval. However, they do not consider the query-target relationship to train the textual inversion network to acquire information for retrieval. In this paper, we propose a novel zero-shot CIR method that is trained end-to-end using masked image-text pairs. By exploiting the abundant image-text pairs that are convenient to obtain with a masking strategy for learning the query-target relationship, it is expected that accurate zero-shot CIR using a retrieval-focused textual inversion network can be realized. Experimental results show the effectiveness of the proposed method., Comment: Accepted as a conference paper in IEEE ICIP 2024

This paper proposes a novel framework to reinforce classification models using language-guided generated counterfactual images. Deep learning classification models are often trained using datasets that mirror real-world scenarios. In this training process, because learning is based solely on correlations with labels, there is a risk that models may learn spurious relationships, such as an overreliance on features not central to the subject, like background elements in images. However, due to the black-box nature of the decision-making process in deep learning models, identifying and addressing these vulnerabilities has been particularly challenging. We introduce a novel framework for reinforcing the classification models, which consists of a two-stage process. First, we identify model weaknesses by testing the model using the counterfactual image dataset, which is generated by perturbed image captions. Subsequently, we employ the counterfactual images as an augmented dataset to fine-tune and reinforce the classification model. Through extensive experiments on several classification models across various datasets, we revealed that fine-tuning with a small set of counterfactual images effectively strengthens the model., Comment: 6 pages, 4 figures

High-momentum two-particle correlations are a useful tool for studying jet-quenching effects in the quark-gluon plasma. Angular correlations between neutral-pion triggers and charged hadrons with transverse momenta in the range 4--12~GeV/$c$ and 0.5--7~GeV/$c$, respectively, have been measured by the PHENIX experiment in 2014 for Au$+$Au collisions at $\sqrt{s_{_{NN}}}=200$~GeV. Suppression is observed in the yield of high-momentum jet fragments opposite the trigger particle, which indicates jet suppression stemming from in-medium partonic energy loss, while enhancement is observed for low-momentum particles. The ratio and differences between the yield in Au$+$Au collisions and $p$$+$$p$ collisions, $I_{AA}$ and $\Delta_{AA}$, as a function of the trigger-hadron azimuthal separation, $\Delta\phi$, are measured for the first time at the Relativistic Heavy Ion Collider. These results better quantify how the yield of low-$p_T$ associated hadrons is enhanced at wide angle, which is crucial for studying energy loss as well as medium-response effects., Comment: 535 authors from 84 institutions, 12 pages, 8 figures. v2 is version accepted for publication in Physical Review C. HEPdata tables for the points plotted in figures for this and previous PHENIX publications are (or will be) publicly available at http://www.phenix.bnl.gov/papers.html

Large language model systems face important security risks from maliciously crafted messages that aim to overwrite the system's original instructions or leak private data. To study this problem, we organized a capture-the-flag competition at IEEE SaTML 2024, where the flag is a secret string in the LLM system prompt. The competition was organized in two phases. In the first phase, teams developed defenses to prevent the model from leaking the secret. During the second phase, teams were challenged to extract the secrets hidden for defenses proposed by the other teams. This report summarizes the main insights from the competition. Notably, we found that all defenses were bypassed at least once, highlighting the difficulty of designing a successful defense and the necessity for additional research to protect LLM systems. To foster future research in this direction, we compiled a dataset with over 137k multi-turn attack chats and open-sourced the platform.

Monitor of All-sky X-ray Image (MAXI) detected a superflare, releasing $5\times 10^{37}$ erg in 2$-$10 keV, of the RS CVn-type star IM Peg at 10:41 UT on 2023 July 23 with its Gas Slit Camera (GSC; 2$-$30 keV). We conducted X-ray follow-up observations of the superflare with Neutron Star Interior Composition ExploreR (NICER; 0.2$-$12 keV) starting at 16:52 UT on July 23 until 06:00 UT on August 2. NICER X-ray spectra clearly showed emission lines of the Fe XXV He$\alpha$ and Fe XXVI Ly$\alpha$ for $\sim 1.5$ days since the MAXI detection. The Fe XXV He$\alpha$ line was blue-shifted with its maximum Doppler velocity reaching $-2200 \pm 600$ $\mathrm{km \: s^{-1}}$, suggesting an upward-moving plasma during the flare, such as a coronal mass ejection (CME) and/or chromospheric evaporation. This is the first case that the Fe XXV He$\alpha$ line is blue-shifted during a stellar flare and its velocity overwhelmingly exceeds the escape velocity of the star ($-230$ $\mathrm{km \: s^{-1}}$). One hour before the most pronounced blueshift detection, a signature of reheating the flare plasma was observed. We discuss the origin of the blueshift, a CME or high-velocity chromospheric evaporation., Comment: 14 pages, 9 figures. Accepted for publication in The Astrophysical Journal Letters