Search

In this technical report, we describe our submission to DCASE2024 Challenge Task6 (Automated Audio Captioning) and Task8 (Language-based Audio Retrieval). We develop our approach building upon the EnCLAP audio captioning framework and optimizing it for Task6 of the challenge. Notably, we outline the changes in the underlying components and the incorporation of the reranking process. Additionally, we submit a supplementary retriever model, a byproduct of our modified framework, to Task8. Our proposed systems achieve FENSE score of 0.542 on Task6 and mAP@10 score of 0.386 on Task8, significantly outperforming the baseline models., Comment: DCASE2024 Challenge Technical Report. Ranked 2nd in Task 6 Automated Audio Captioning

Driven-dissipative nonlinear systems exhibit rich critical behavior, related to bifurcation, bistability and switching, which underlie key phenomena in areas ranging from physics, chemistry and biology to social sciences and economics. The importance of rare fluctuations leading to a dramatic jump between two very distinct states, such as survival and extinction in population dynamics, success and bankruptcy in economics and the occurrence of earthquakes or of epileptic seizures, have been already established. In the quantum domain, switching is of importance in both chemical reactions and the devices used in quantum state detection and amplification. In particular, the simplest driven single oscillator model serves as an insightful starting point. Here we describe switching induced by quantum fluctuations and illustrate that an instanton approach within Keldysh field theory can provide a deep insight into such phenomena. We provide a practical recipe to compute the switching rates semi-analytically, which agrees remarkably well with exact solutions across a wide domain of drive amplitudes spanning many orders of magnitude. Being set up in the framework of Keldysh coherent states path integrals, our approach opens the possibility of studying quantum activation in many-body systems where other approaches are inapplicable.

We explore the higher integrability of Green's functions associated with the second-order elliptic equation $a^{ij}D_{ij}u + b^i D_iu = f$ in a bounded domain $\Omega \subset \mathbb{R}^d$, and establish a version of Aleksandrov's maximum principle. In particular, we consider the drift term $b=(b^1, \ldots, b^d)$ in $L_d$ and the source term $f \in L_p$ for some $p < d$. This provides an alternative and analytic proof of a result by N. V. Krylov (2021) concerning $L_d$ drifts. The key step involves deriving a Gehring type inequality for Green's functions by using the Fabes-Stroock approach (1984).

In this work, we consider the problem of training a generator from evaluations of energy functions or unnormalized densities. This is a fundamental problem in probabilistic inference, which is crucial for scientific applications such as learning the 3D coordinate distribution of a molecule. To solve this problem, we propose iterated energy-based flow matching (iEFM), the first off-policy approach to train continuous normalizing flow (CNF) models from unnormalized densities. We introduce the simulation-free energy-based flow matching objective, which trains the model to predict the Monte Carlo estimation of the marginal vector field constructed from known energy functions. Our framework is general and can be extended to variance-exploding (VE) and optimal transport (OT) conditional probability paths. We evaluate iEFM on a two-dimensional Gaussian mixture model (GMM) and an eight-dimensional four-particle double-well potential (DW-4) energy function. Our results demonstrate that iEFM outperforms existing methods, showcasing its potential for efficient and scalable probabilistic modeling in complex high-dimensional systems.

Reinforcement learning (RL) with diverse offline datasets can have the advantage of leveraging the relation of multiple tasks and the common skills learned across those tasks, hence allowing us to deal with real-world complex problems efficiently in a data-driven way. In offline RL where only offline data is used and online interaction with the environment is restricted, it is yet difficult to achieve the optimal policy for multiple tasks, especially when the data quality varies for the tasks. In this paper, we present a skill-based multi-task RL technique on heterogeneous datasets that are generated by behavior policies of different quality. To learn the shareable knowledge across those datasets effectively, we employ a task decomposition method for which common skills are jointly learned and used as guidance to reformulate a task in shared and achievable subtasks. In this joint learning, we use Wasserstein auto-encoder (WAE) to represent both skills and tasks on the same latent space and use the quality-weighted loss as a regularization term to induce tasks to be decomposed into subtasks that are more consistent with high-quality skills than others. To improve the performance of offline RL agents learned on the latent space, we also augment datasets with imaginary trajectories relevant to high-quality skills for each task. Through experiments, we show that our multi-task offline RL approach is robust to the mixed configurations of different-quality datasets and it outperforms other state-of-the-art algorithms for several robotic manipulation tasks and drone navigation tasks., Comment: 12 pages, 5 figures, acceepted in NeurIPS 2022

We revisit the relation between active galactic nucleus (AGN) broad-line region (BLR) size and luminosity by conducting a uniform H$\beta$ reverberation-mapping analysis for 212 AGNs with archival light curves. Our analysis incorporates three different lag measurement methods, including the interpolated cross-correlation function (ICCF), JAVELIN, and PyROA, alongside a consistently defined lag searching window and an alias removal procedure. We find that ICCF, albeit with larger uncertainties compared to other methods, is the most reliable method based on our visual inspection of the matches between H$\beta$ and the shifted continuum light curves. Combining this sample with the 32 AGNs from the Seoul National University AGN Monitoring Project, we obtain the best-fit relation between the BLR size ($R_{\rm BLR}$) and the continuum luminosity at 5100$\r{A}$ ($L_{5100}$) with a slope significantly flatter than 0.5. By selecting a subsample of 157 AGNs with the best-quality lag measurements using a set of quantitative criteria and visual inspection, we find a consistent slope and a slightly decreased intrinsic scatter. We further investigate the effect of luminosity tracers, including $L_{5100}$, H$\beta$ luminosity ($L_{\rm H\beta}$), ${\rm [O\,\mathrm{III}]}$ luminosity ($L_{\rm [O\,\mathrm{III}]}$), and 2 to 10 keV hard X-ray luminosity ($L_{\rm 2\unicode{x2013}10\,keV}$). We find that sub-Eddington and super-Eddington AGNs exhibit systematic offsets in both $R_{\rm BLR}$$\unicode{x2013}$$L_{5100}$ and $R_{\rm BLR}$$\unicode{x2013}$$L_{\rm H\beta}$ relation, but are comparable in the $R_{\rm BLR}$$\unicode{x2013}$$L_{\rm [O\,\mathrm{III}]}$ and $R_{\rm BLR}$$\unicode{x2013}$$L_{\rm 2\unicode{x2013}10\,keV}$ relation. We discuss the potential causes for these different deviations when employing different luminosity tracers., Comment: 20 pages, 9 figures; accepted for publication in the Astrophysical Journal Supplement Series

A high-brightness entangled photon pair (HBEPP) source is essential for conducting entanglement-based quantum key distribution (QKD) between a satellite and a ground station. While an ultrabright source can overcome significant losses in satellite-based QKD (SQKD) and increase the sifted key rate, it also induces the multi-photon effect, raising the system's error rate. To accurately estimate system performance, we first present an analytical model for calculating the measurement probabilities of HBEPP distribution through an asymmetric loss channel. Based on this model, we propose the use of a passive-intensity HBEPP source for SQKD systems, assuming a polarization-independent channel and threshold detectors for measurement. We confirm that fixing the mean photon number at $\bar{\mu}=0.1$ achieves a performance of $99.7\%$ compared to the ideal one-way communication entanglement-based SQKD protocol, which is effectively optimizing the HBEPP source brightness in accordance with system losses., Comment: 9 pages, 5 figures, 3 tables

Kramers-Kronig (KK) analysis has been widely used to extract the optical conductivity spectrum from a broad range of reflectance spectrum obtained from far-infrared to ultraviolet frequency ranges. In this study, we present how measurement uncertainty in the reflectivity spectrum affects the extracted optical conductivity spectrum obtained through KK analysis. We consider realistic uncertainties that can easily occur in reflectance measurement environments: (1) a rigid shift of the absolute reflectance in the whole measurement frequency window, and (2) a linear decrement of reflectance with increasing frequency. Our investigation reveals that the reliability of the extracted optical conductivity spectrum, especially in the lower-frequency range, should be carefully addressed, particularly when the reflectance is above approximately 95 %., Comment: 7 pages with 3 figures

Quantum error correction provides a path to reach practical quantum computing by combining multiple physical qubits into a logical qubit, where the logical error rate is suppressed exponentially as more qubits are added. However, this exponential suppression only occurs if the physical error rate is below a critical threshold. In this work, we present two surface code memories operating below this threshold: a distance-7 code and a distance-5 code integrated with a real-time decoder. The logical error rate of our larger quantum memory is suppressed by a factor of $\Lambda$ = 2.14 $\pm$ 0.02 when increasing the code distance by two, culminating in a 101-qubit distance-7 code with 0.143% $\pm$ 0.003% error per cycle of error correction. This logical memory is also beyond break-even, exceeding its best physical qubit's lifetime by a factor of 2.4 $\pm$ 0.3. We maintain below-threshold performance when decoding in real time, achieving an average decoder latency of 63 $\mu$s at distance-5 up to a million cycles, with a cycle time of 1.1 $\mu$s. To probe the limits of our error-correction performance, we run repetition codes up to distance-29 and find that logical performance is limited by rare correlated error events occurring approximately once every hour, or 3 $\times$ 10$^9$ cycles. Our results present device performance that, if scaled, could realize the operational requirements of large scale fault-tolerant quantum algorithms., Comment: 10 pages, 4 figures, Supplementary Information

The Berry curvature dipole (BCD) serves as a one of the fundamental contributors to emergence of the nonlinear Hall effect (NLHE). Despite intense interest due to its potential for new technologies reaching beyond the quantum efficiency limit, the interplay between BCD and NLHE has been barely understood yet in the absence of a systematic study on the electronic band structure. Here, we report NLHE realized in NbIrTe4 that persists above room temperature coupled with a sign change in the Hall conductivity at 150 K. First-principles calculations combined with angle-resolved photoemission spectroscopy (ARPES) measurements show that BCD tuned by the partial occupancy of spin-orbit split bands via temperature is responsible for the temperature-dependent NLHE. Our findings highlight the correlation between BCD and the electronic band structure, providing a viable route to create and engineer the non-trivial Hall effect by tuning the geometric properties of quasiparticles in transition-metal chalcogen compounds.

Goal-conditioned (GC) policy learning often faces a challenge arising from the sparsity of rewards, when confronting long-horizon goals. To address the challenge, we explore skill-based GC policy learning in offline settings, where skills are acquired from existing data and long-horizon goals are decomposed into sequences of near-term goals that align with these skills. Specifically, we present an `offline GC policy learning via skill-step abstraction' framework (GLvSA) tailored for tackling long-horizon GC tasks affected by goal distribution shifts. In the framework, a GC policy is progressively learned offline in conjunction with the incremental modeling of skill-step abstractions on the data. We also devise a GC policy hierarchy that not only accelerates GC policy learning within the framework but also allows for parameter-efficient fine-tuning of the policy. Through experiments with the maze and Franka kitchen environments, we demonstrate the superiority and efficiency of our GLvSA framework in adapting GC policies to a wide range of long-horizon goals. The framework achieves competitive zero-shot and few-shot adaptation performance, outperforming existing GC policy learning and skill-based methods., Comment: 9 pages, 4 figures, International Joint Conference on Artificial Intelligence 2024, Published version

Vanadium-based kagome systems AV$_3$Sb$_5$ (A = K, Rb, Cs) have emerged as paradigmatic examples exhibiting unconventional charge density waves (CDWs) and superconductivity linked to van Hove singularities (VHSs). Despite extensive studies, the three-dimensional (3D) nature of CDW states in these systems remains elusive. This study employs first-principles density functional theory and a tight-binding model to investigate the stacking-dependent electronic structures of 3D CDWs in AV$_3$Sb$_5$, emphasizing the significant role of interlayer coupling in behaviors of the VHSs associated with diverse 3D CDW orders. We develop a minimal 3D tight-binding model and present a detailed analysis of band structures and density of states for various 3D CDW stacking configurations, including those with and without a $\pi$-phase shift stacking of the inverse star of David, as well as alternating stacking of the inverse star of David and the star of David. We find that VHSs exist below the Fermi level even in 3D CDWs without $\pi$-phase shift stackings, and that these VHSs shift downward in the $\pi$-phase shift stacking CDW structure, stabilizing the $2\times2\times2$ $\pi$-shifted inverse star of David distortions in alternating vanadium layers as the ground state 3D CDW order of AV$_3$Sb$_5$. Our work provides the electronic origin of 3D CDW orders, paving the way for a deeper understanding of CDWs and superconductivity in AV$_3$Sb$_5$ kagome metals., Comment: 4 figures + Supplementary information

Open-vocabulary 3D instance segmentation transcends traditional closed-vocabulary methods by enabling the identification of both previously seen and unseen objects in real-world scenarios. It leverages a dual-modality approach, utilizing both 3D point clouds and 2D multi-view images to generate class-agnostic object mask proposals. Previous efforts predominantly focused on enhancing 3D mask proposal models; consequently, the information that could come from 2D association to 3D was not fully exploited. This bias towards 3D data, while effective for familiar indoor objects, limits the system's adaptability to new and varied object types, where 2D models offer greater utility. Addressing this gap, we introduce Zero-Shot Dual-Path Integration Framework that equally values the contributions of both 3D and 2D modalities. Our framework comprises three components: 3D pathway, 2D pathway, and Dual-Path Integration. 3D pathway generates spatially accurate class-agnostic mask proposals of common indoor objects from 3D point cloud data using a pre-trained 3D model, while 2D pathway utilizes pre-trained open-vocabulary instance segmentation model to identify a diverse array of object proposals from multi-view RGB-D images. In Dual-Path Integration, our Conditional Integration process, which operates in two stages, filters and merges the proposals from both pathways adaptively. This process harmonizes output proposals to enhance segmentation capabilities. Our framework, utilizing pre-trained models in a zero-shot manner, is model-agnostic and demonstrates superior performance on both seen and unseen data, as evidenced by comprehensive evaluations on the ScanNet200 and qualitative results on ARKitScenes datasets., Comment: OpenSUN 3D: 2nd Workshop on Open-Vocabulary 3D Scene Understanding (CVPR 2024)

Understanding traffic behavior is crucial for enhancing the stable functioning and safety of transportation systems. Previous percolation-based transportation studies have analyzed transition behaviors from free-flow to traffic-jam states, with a focus on robustness and resilience during congestion. However, relatively less attention is paid to the percolation analysis of the free-flow states, specifically how free-flow clusters form and grow. In this study, we investigate the percolation patterns of two opposing traffic scenarios -- traffic jam state and free-flow state -- within the same road network using Chengdu taxi data and compare their percolating behaviors. Our analysis reveals differences between the two scenarios in the growth patterns of the giant connected component (GCC), which is captured by a persistent gap between the GCC size curves, particularly during peak hours. We attribute these disparities to a long-range spatial correlation of traffic speed within a road network. Empirically, we find distinct long-range spatial correlations in traffic, using rescaled taxi speeds on roads, and we examine their relationship with each percolation pattern. Our analysis provides an integrated view of traffic dynamics and uncovers intrinsic traffic correlations within urban areas that drive these intriguing percolation patterns. Our findings also offer valuable metrics for effective traffic management and accident prevention strategies, aligning with urban transportation safety and reliability goals. These insights are beneficial for assessing and designing resilient urban road networks that maintain functionality under stress, ultimately improving the reliability of traffic assessments and reducing accidents., Comment: 17 pages, 10 figures

Complex urban road networks with high vehicle occupancy frequently face severe traffic congestion. Designing an effective strategy for managing multiple traffic lights plays a crucial role in managing congestion. However, most current traffic light management systems rely on human-crafted decisions, which may not adapt well to diverse traffic patterns. In this paper, we delve into two pivotal design components of the traffic light management system that can be dynamically adjusted to various traffic conditions: phase combination and phase time allocation. While numerous studies have sought an efficient strategy for managing traffic lights, most of these approaches consider a fixed traffic pattern and are limited to relatively small road networks. To overcome these limitations, we introduce a novel and practical framework to formulate the optimization of such design components using an offline meta black-box optimization. We then present a simple yet effective method to efficiently find a solution for the aforementioned problem. In our framework, we first collect an offline meta dataset consisting of pairs of design choices and corresponding congestion measures from various traffic patterns. After collecting the dataset, we employ the Attentive Neural Process (ANP) to predict the impact of the proposed design on congestion across various traffic patterns with well-calibrated uncertainty. Finally, Bayesian optimization, with ANP as a surrogate model, is utilized to find an optimal design for unseen traffic patterns through limited online simulations. Our experiment results show that our method outperforms state-of-the-art baselines on complex road networks in terms of the number of waiting vehicles. Surprisingly, the deployment of our method into a real-world traffic system was able to improve traffic throughput by 4.80\% compared to the original strategy., Comment: 12 pages, 7 figures, 10 tables

Ventilated acoustic resonator(VAR), a type of acoustic metamaterial, emerge as an alternative for sound attenuation in environments that require ventilation, owing to its excellent low-frequency attenuation performance and flexible shape adaptability. However, due to the non-linear acoustic responses of VARs, the VAR designs are generally obtained within a limited parametrized design space, and the design relies on the iteration of the numerical simulation which consumes a considerable amount of computational time and resources. This paper proposes an acoustic response-encoded variational autoencoder (AR-VAE), a novel variational autoencoder-based generative design model for the efficient and accurate inverse design of VAR even with non-parametrized designs. The AR-VAE matches the high-dimensional acoustic response with the VAR cross-section image in the dimension-reduced latent space, which enables the AR-VAE to generate various non-parametrized VAR cross-section images with the target acoustic response. AR-VAE generates non-parameterized VARs from target acoustic responses, which show a 25-fold reduction in mean squared error compared to conventional deep learning-based parameter searching methods while exhibiting lower average mean squared error and peak frequency variance. By combining the inverse-designed VARs by AR-VAE, multi-cavity VAR was devised for broadband and multitarget peak frequency attenuation. The proposed design method presents a new approach for structural inverse-design with a high-dimensional non-linear physical response.

We present Blind-Match, a novel biometric identification system that leverages homomorphic encryption (HE) for efficient and privacy-preserving 1:N matching. Blind-Match introduces a HE-optimized cosine similarity computation method, where the key idea is to divide the feature vector into smaller parts for processing rather than computing the entire vector at once. By optimizing the number of these parts, Blind-Match minimizes execution time while ensuring data privacy through HE. Blind-Match achieves superior performance compared to state-of-the-art methods across various biometric datasets. On the LFW face dataset, Blind-Match attains a 99.63% Rank-1 accuracy with a 128-dimensional feature vector, demonstrating its robustness in face recognition tasks. For fingerprint identification, Blind-Match achieves a remarkable 99.55% Rank-1 accuracy on the PolyU dataset, even with a compact 16-dimensional feature vector, significantly outperforming the state-of-the-art method, Blind-Touch, which achieves only 59.17%. Furthermore, Blind-Match showcases practical efficiency in large-scale biometric identification scenarios, such as Naver Cloud's FaceSign, by processing 6,144 biometric samples in 0.74 seconds using a 128-dimensional feature vector., Comment: Accepted to CIKM 2024 (Applied Research Track)

The electroweak monopole, when coupled to gravity, turns to the Reissner-Nordstrom type primordial magnetic blackhole whose mass is bounded below, with the lower bound $M_P \sqrt \alpha$. This changes the overall picture of the monopole production mechanism in the early universe drastically and has deep implications in cosmolpgy. In particular, this enhances the possibility that the electroweak monopoles turned to the primordial magnetic blackholes could become the seed of stellar objects and galaxies, and account for the dark matter of the universe. Moreover, this tells that we have a new type of primordial blackhole different from the popular primordial blackhole in cosmology, the electroweak primordial magnetic blackhole based on a totally different production mechanism. We discuss the physical implications of the electroweak primordial magnetic blackhole.

Copper-doped lead apatite, called LK-99, was initially claimed to be a room temperature superconductor driven by flat electron bands, but was later found to be a wide gap insulator. Despite the lack of room temperature superconductivity, there is growing evidence that LK-99 and related compounds host various strong electron correlation phenomena arising from their flat electron bands. Depending on the copper doping site and crystal structure, LK-99 can exhibit two distinct flat bands crossing the Fermi level in the non-interacting limit: either a single or two entangled flat bands. We explore potential correlated metallic and insulating phases in the flat bands of LK-99 compounds by constructing their correlation phase diagrams, and find both non-Fermi liquid and Mott insulating states. We demonstrate that LK-99 is a charge-transfer Mott insulator driven by strong electron correlations, regardless of the flat band type. We also find that the non-Fermi liquid state in the multi-flat band system exhibits strange metal behaviour, while the corresponding state in the single flat band system exhibits pseudogap behaviour. Our findings align with available experimental observations and provide crucial insights into the correlation phenomenology of LK-99 and related compounds that could arise independently of superconductivity. Overall, our research highlights that LK-99 and related compounds offer a compelling platform for investigating correlation physics in flat band systems., Comment: 5 figures + Supplementary information

Accurate survival prediction in radiotherapy (RT) is critical for optimizing treatment decisions. This study developed and validated the RT-Surv framework, which integrates general-domain, open-source large language models (LLMs) to structure unstructured electronic health records alongside structured clinical data. Using data from 34,276 patients and an external cohort of 852, the framework successfully transformed unstructured clinical information into structured formats. Incorporating LLM-structured clinical features improved the concordance index from 0.779 to 0.842 during external validation, demonstrating a significant performance enhancement. Key LLM-structured features, such as disease extent, general condition, and RT purpose, showed high predictive importance and aligned closely with statistically significant predictors identified through conventional statistical analyses, thereby improving model interpretability. Furthermore, the framework enhanced risk stratification, enabling more distinct differentiation among low-, intermediate-, and high-risk groups (p < 0.001) using LLM-structured clinical features. These findings highlight the potential of LLMs to convert unstructured data into actionable insights, improving predictive modeling and patient outcomes in clinics., Comment: 23 pages, 2 tables, 4 figures

The geometric characteristics of Bloch wave functions play a crucial role in electronic transport properties. We show that the thermoelectric performance of materials is governed by the geometric structure of Bloch wave functions within the framework of the Boltzmann equation. The essential geometric notion is the Hilbert-Schmidt quantum distance, measuring the resemblance between two quantum states. We establish a geometric characterization of the scattering rate by extending the concept of quantum distance between two states in momentum space at a distance.Employing isotropic quadratic band touching semimetals, where one can concentrate on the role of quantum geometric effects other than the Berry curvature, we find that the response functions for electrical quantum transport and, therefore, the thermoelectric power factor can be succinctly expressed in terms of the maximum quantum distance, $d_\mathrm{max}$. Specifically, when $d_\mathrm{max}$ reaches one, the power factor doubles compared to the case with trivial geometry ($d_\mathrm{max}=0$). Our finding highlights the significance of quantum geometry in improving the performance of thermoelectric devices.

Retrieval-Augmented Language Models (RALMs) have significantly improved performance in open-domain question answering (QA) by leveraging external knowledge. However, RALMs still struggle with unanswerable queries, where the retrieved contexts do not contain the correct answer, and with conflicting information, where different sources provide contradictory answers due to imperfect retrieval. This study introduces an in-context learning-based approach to enhance the reasoning capabilities of RALMs, making them more robust in imperfect retrieval scenarios. Our method incorporates Machine Reading Comprehension (MRC) demonstrations, referred to as cases, to boost the model's capabilities to identify unanswerabilities and conflicts among the retrieved contexts. Experiments on two open-domain QA datasets show that our approach increases accuracy in identifying unanswerable and conflicting scenarios without requiring additional fine-tuning. This work demonstrates that in-context learning can effectively enhance the robustness of RALMs in open-domain QA tasks., Comment: 10 pages, 2 figures

To enable remote Virtual Reality (VR) and Augmented Reality (AR) clients to collaborate as if they were in the same space during Mixed Reality (MR) telepresence, it is essential to overcome spatial heterogeneity and generate a unified shared collaborative environment by integrating remote spaces into a target host space. Especially when multiple remote users connect, a large shared space is necessary for people to maintain their personal space while collaborating, but the existing simple intersection method leads to the creation of narrow shared spaces as the number of remote spaces increases. To robustly align to the host space even as the number of remote spaces increases, we propose a spatial affordance-aware interactable subspace allocation algorithm. The key concept of our approach is to consider the perceivable and interactable areas separately, where every user views the same mutual space, but each remote user has a different interactable subspace, considering their location and spatial affordance. We conducted an evaluation with 900 space combinations, varying the number of remote spaces as two, four, and six, and results show our method outperformed in securing wide interactable mutual space and instantiating users compared to the other spatial matching methods. Our work enables multiple clients from diverse remote locations to access the AR host's space, allowing them to interact directly with the table, wall, or floor by aligning their physical subspaces within a connected mutual space., Comment: Accepted at the 2024 IEEE ISMAR Conference. 10 pages, 6 figures

Fully supervised training of semantic segmentation models is costly and challenging because each pixel within an image needs to be labeled. Therefore, the sparse pixel-level annotation methods have been introduced to train models with a subset of pixels within each image. We introduce a Bayesian active learning framework based on sparse pixel-level annotation that utilizes a pixel-level Bayesian uncertainty measure based on Balanced Entropy (BalEnt) [84]. BalEnt captures the information between the models' predicted marginalized probability distribution and the pixel labels. BalEnt has linear scalability with a closed analytical form and can be calculated independently per pixel without relational computations with other pixels. We train our proposed active learning framework for Cityscapes, Camvid, ADE20K and VOC2012 benchmark datasets and show that it reaches supervised levels of mIoU using only a fraction of labeled pixels while outperforming the previous state-of-the-art active learning models with a large margin.

Quantum conductors attached to metallic reservoirs have been demonstrated to overcome the thermodynamic uncertainty relation (TUR), a trade-off relation between the amount of dissipation and the absence of charge and heat current fluctuations. Here, we report large TUR violations when superconducting reservoirs replace metallic ones. The coexistence of different transport processes, namely (multiple) Andreev reflection, where electrons and their retro-reflected holes create Cooper pairs, in addition to the normal quasiparticle transport is identified as the source for such TUR breakdowns. The large TUR violation is a remarkable advantage for building low dissipative and highly stable quantum thermal machines., Comment: 6 pages, 2 figures

We propose a fault-tolerant quantum computation scheme in a measurement-based manner with finite-sized entangled resource states and encoded fusion scheme with linear optics. The encoded-fusion is an entangled measurement devised to enhance the fusion success probability in the presence of losses and errors based on a quantum error-correcting code. We apply an encoded-fusion scheme, which can be performed with linear optics and active feedforwards to implement the generalized Shor code, to construct a fault-tolerant network configuration in a three-dimensional Raussendorf-Harrington-Goyal lattice based on the surface code. Numerical simulations show that our scheme allows us to achieve up to 10 times higher loss thresholds than nonencoded fusion approaches with limited numbers of photons used in fusion. Our scheme paves an efficient route toward fault-tolerant quantum computing with finite-sized entangled resource states and linear optics., Comment: 14 pages including 5 Appendices, 14 figures

In embodied instruction-following (EIF), the integration of pretrained language models (LMs) as task planners emerges as a significant branch, where tasks are planned at the skill level by prompting LMs with pretrained skills and user instructions. However, grounding these pretrained skills in different domains remains challenging due to their intricate entanglement with the domain-specific knowledge. To address this challenge, we present a semantic skill grounding (SemGro) framework that leverages the hierarchical nature of semantic skills. SemGro recognizes the broad spectrum of these skills, ranging from short-horizon low-semantic skills that are universally applicable across domains to long-horizon rich-semantic skills that are highly specialized and tailored for particular domains. The framework employs an iterative skill decomposition approach, starting from the higher levels of semantic skill hierarchy and then moving downwards, so as to ground each planned skill to an executable level within the target domain. To do so, we use the reasoning capabilities of LMs for composing and decomposing semantic skills, as well as their multi-modal extension for assessing the skill feasibility in the target domain. Our experiments in the VirtualHome benchmark show the efficacy of SemGro in 300 cross-domain EIF scenarios., Comment: Findings of ACL-2024 Camera Ready Version

Delineating lesions and anatomical structure is important for image-guided interventions. Point-supervised medical image segmentation (PSS) has great potential to alleviate costly expert delineation labeling. However, due to the lack of precise size and boundary guidance, the effectiveness of PSS often falls short of expectations. Although recent vision foundational models, such as the medical segment anything model (MedSAM), have made significant advancements in bounding-box-prompted segmentation, it is not straightforward to utilize point annotation, and is prone to semantic ambiguity. In this preliminary study, we introduce an iterative framework to facilitate semantic-aware point-supervised MedSAM. Specifically, the semantic box-prompt generator (SBPG) module has the capacity to convert the point input into potential pseudo bounding box suggestions, which are explicitly refined by the prototype-based semantic similarity. This is then succeeded by a prompt-guided spatial refinement (PGSR) module that harnesses the exceptional generalizability of MedSAM to infer the segmentation mask, which also updates the box proposal seed in SBPG. Performance can be progressively improved with adequate iterations. We conducted an evaluation on BraTS2018 for the segmentation of whole brain tumors and demonstrated its superior performance compared to traditional PSS methods and on par with box-supervised methods., Comment: 2024 IEEE Nuclear Science Symposium and Medical Imaging Conference

The concept of Chern insulators is one of the most important buliding block of topological physics, enabling the quantum Hall effect without external magnetic fields. The construction of Chern insulators has been typically through an guess-and-confirm approach, which can be inefficient and unpredictable. In this paper, we introduce a systematic method to directly construct two-dimensional Chern insulators that can provide any nontrivial Chern number. Our method is built upon the one-dimensional Rice-Mele model, which is well known for its adjustable polarization properties, providing a reliable framework for manipulation. By extending this model into two dimensions, we are able to engineer lattice structures that demonstrate predetermined topological quantities effectively. This research not only contributes the development of Chern insulators but also paves the way for designing a variety of lattice structures with significant topological implications, potentially impacting quantum computing and materials science. With this approach, we are to shed light on the pathways for designing more complex and functional topological phases in synthetic materials., Comment: 18 pages, 6 figures

Self-supervised multi-frame monocular depth estimation relies on the geometric consistency between successive frames under the assumption of a static scene. However, the presence of moving objects in dynamic scenes introduces inevitable inconsistencies, causing misaligned multi-frame feature matching and misleading self-supervision during training. In this paper, we propose a novel framework called ProDepth, which effectively addresses the mismatch problem caused by dynamic objects using a probabilistic approach. We initially deduce the uncertainty associated with static scene assumption by adopting an auxiliary decoder. This decoder analyzes inconsistencies embedded in the cost volume, inferring the probability of areas being dynamic. We then directly rectify the erroneous cost volume for dynamic areas through a Probabilistic Cost Volume Modulation (PCVM) module. Specifically, we derive probability distributions of depth candidates from both single-frame and multi-frame cues, modulating the cost volume by adaptively fusing those distributions based on the inferred uncertainty. Additionally, we present a self-supervision loss reweighting strategy that not only masks out incorrect supervision with high uncertainty but also mitigates the risks in remaining possible dynamic areas in accordance with the probability. Our proposed method excels over state-of-the-art approaches in all metrics on both Cityscapes and KITTI datasets, and demonstrates superior generalization ability on the Waymo Open dataset., Comment: Accepted by ECCV 2024. Project Page: https://sungmin-woo.github.io/prodepth/

Magnetic anisotropy in atomically thin correlated heterostructures is essential for exploring quantum magnetic phases for next-generation spintronics. Whereas previous studies have mostly focused on van der Waals systems, here, we investigate the impact of dimensionality of epitaxially-grown correlated oxides down to the monolayer limit on structural, magnetic, and orbital anisotropies. By designing oxide superlattices with a correlated ferromagnetic SrRuO3 and nonmagnetic SrTiO3 layers, we observed modulated ferromagnetic behavior with the change of the SrRuO3 thickness. Especially, for three-unit-cell-thick layers, we observe a significant 1,500% improvement of coercive field in the anomalous Hall effect, which cannot be solely attributed to the dimensional crossover in ferromagnetism. The atomic-scale heterostructures further reveal the systematic modulation of anisotropy for the lattice structure and orbital hybridization, explaining the enhanced magnetic anisotropy. Our findings provide valuable insights into engineering the anisotropic hybridization of synthetic magnetic crystals, offering a tunable spin order for various applications., Comment: 23 pages

In this paper, we propose an algorithm to generate a static point cloud map based on LiDAR point cloud data. Our proposed pipeline detects dynamic objects using 3D object detectors and projects points of dynamic objects onto the ground. Typically, point cloud data acquired in real-time serves as a snapshot of the surrounding areas containing both static objects and dynamic objects. The static objects include buildings and trees, otherwise, the dynamic objects contain objects such as parked cars that change their position over time. Removing dynamic objects from the point cloud map is crucial as they can degrade the quality and localization accuracy of the map. To address this issue, in this paper, we propose an algorithm that creates a map only consisting of static objects. We apply a 3D object detection algorithm to the point cloud data which are obtained from LiDAR to implement our pipeline. We then stack the points to create the map after performing ground segmentation and projection. As a result, not only we can eliminate currently dynamic objects at the time of map generation but also potentially dynamic objects such as parked vehicles. We validate the performance of our method using two kinds of datasets collected on real roads: KITTI and our dataset. The result demonstrates the capability of our proposal to create an accurate static map excluding dynamic objects from input point clouds. Also, we verified the improved performance of localization using a generated map based on our method.

We present experimental measurements on Young's modulus and Gr\"{u}neisen parameter of multilayer graphene with varying number of layers using in situ bulge tests corroborated by atomic force microscopy and Raman spectroscopy. Due to the experimental challenges posed by the significant disparity between intra and interlayer mechanical strengths, measuring conclusive elastic parameters are proven difficult. We analyze the results based on a new theoretical model together with first-principles calculations to elucidate the mechanism of thickness dependent incomplete strain transfer across the layers. Our findings indicate that the experimental elastic constants, which differ from intrinsic computed ones, have unavoidable limitations in reflecting ideal mechanical couplings between layers. Consequently, we propose a new mechanical modulus, free from effects of uncontrollable delamination of layers, derived by combining Gr\"{u}neisen parameter with Young's modulus., Comment: 6 pages, 4 figures

Large language models (LLMs) have made significant advancements in various natural language processing tasks, including question answering (QA) tasks. While incorporating new information with the retrieval of relevant passages is a promising way to improve QA with LLMs, the existing methods often require additional fine-tuning which becomes infeasible with recent LLMs. Augmenting retrieved passages via prompting has the potential to address this limitation, but this direction has been limitedly explored. To this end, we design a simple yet effective framework to enhance open-domain QA (ODQA) with LLMs, based on the summarized retrieval (SuRe). SuRe helps LLMs predict more accurate answers for a given question, which are well-supported by the summarized retrieval that could be viewed as an explicit rationale extracted from the retrieved passages. Specifically, SuRe first constructs summaries of the retrieved passages for each of the multiple answer candidates. Then, SuRe confirms the most plausible answer from the candidate set by evaluating the validity and ranking of the generated summaries. Experimental results on diverse ODQA benchmarks demonstrate the superiority of SuRe, with improvements of up to 4.6% in exact match (EM) and 4.0% in F1 score over standard prompting approaches. SuRe also can be integrated with a broad range of retrieval methods and LLMs. Finally, the generated summaries from SuRe show additional advantages to measure the importance of retrieved passages and serve as more preferred rationales by models and humans., Comment: Accepted at ICLR 2024