Search

The integration of deep learning technologies in medical imaging aims to enhance the efficiency and accuracy of cancer diagnosis, particularly for pancreatic and breast cancers, which present significant diagnostic challenges due to their high mortality rates and complex imaging characteristics. This paper introduces Diffusion-Driven Diagnosis (D-Cube), a novel approach that leverages hyper-features from a diffusion model combined with contrastive learning to improve cancer diagnosis. D-Cube employs advanced feature selection techniques that utilize the robust representational capabilities of diffusion models, enhancing classification performance on medical datasets under challenging conditions such as data imbalance and limited sample availability. The feature selection process optimizes the extraction of clinically relevant features, significantly improving classification accuracy and demonstrating resilience in imbalanced and limited data scenarios. Experimental results validate the effectiveness of D-Cube across multiple medical imaging modalities, including CT, MRI, and X-ray, showing superior performance compared to existing baseline models. D-Cube represents a new strategy in cancer detection, employing advanced deep learning techniques to achieve state-of-the-art diagnostic accuracy and efficiency., Comment: 10 pages, 2 figures

Refined canonical stable Grothendieck polynomials were introduced by Hwang, Jang, Kim, Song, and Song. There exist two combinatorial models for these polynomials: one using hook-valued tableaux and the other using pairs of a semistandard Young tableau and (what we call) an exquisite tableau. An uncrowding algorithm on hook-valued tableaux was introduced by Pan, Pappe, Poh, and Schilling. In this paper, we discover a novel connection between the two models via the uncrowding and Goulden--Greene's jeu de taquin algorithms, using a classical result of Benkart, Sottile, and Stroomer on tableau switching. This connection reveals a hidden symmetry of the uncrowding algorithm defined on hook-valued tableaux. As a corollary, we obtain another combinatorial model for the refined canonical stable Grothendieck polynomials in terms of biflagged tableaux, which naturally appear in the characterization of the image of the uncrowding map., Comment: 18 pages

This paper introduces ELMO, a real-time upsampling motion capture framework designed for a single LiDAR sensor. Modeled as a conditional autoregressive transformer-based upsampling motion generator, ELMO achieves 60 fps motion capture from a 20 fps LiDAR point cloud sequence. The key feature of ELMO is the coupling of the self-attention mechanism with thoughtfully designed embedding modules for motion and point clouds, significantly elevating the motion quality. To facilitate accurate motion capture, we develop a one-time skeleton calibration model capable of predicting user skeleton offsets from a single-frame point cloud. Additionally, we introduce a novel data augmentation technique utilizing a LiDAR simulator, which enhances global root tracking to improve environmental understanding. To demonstrate the effectiveness of our method, we compare ELMO with state-of-the-art methods in both image-based and point cloud-based motion capture. We further conduct an ablation study to validate our design principles. ELMO's fast inference time makes it well-suited for real-time applications, exemplified in our demo video featuring live streaming and interactive gaming scenarios. Furthermore, we contribute a high-quality LiDAR-mocap synchronized dataset comprising 20 different subjects performing a range of motions, which can serve as a valuable resource for future research. The dataset and evaluation code are available at {\blue \url{https://movin3d.github.io/ELMO_SIGASIA2024/}}, Comment: published at ACM Transactions on Graphics (Proc. SIGGRAPH ASIA), 2024

Recent advancements in large language models (LLMs) have demonstrated impressive performance in generating molecular structures as drug candidates, which offers significant potential to accelerate drug discovery. However, the current LLMs overlook a critical requirement for drug discovery: proposing a diverse set of molecules. This diversity is essential for improving the chances of finding a viable drug, as it provides alternative molecules that may succeed where others fail in wet-lab or clinical validations. Despite such a need for diversity, the LLMs often output structurally similar molecules from a given prompt. While decoding schemes like beam search may enhance textual diversity, this often does not align with molecular structural diversity. In response, we propose a new method for fine-tuning molecular generative LLMs to autoregressively generate a set of structurally diverse molecules, where each molecule is generated by conditioning on the previously generated molecules. Our approach consists of two stages: (1) supervised fine-tuning to adapt LLMs to autoregressively generate molecules in a sequence and (2) reinforcement learning to maximize structural diversity within the generated molecules. Our experiments show that (1) our fine-tuning approach enables the LLMs to better discover diverse molecules compared to existing decoding schemes and (2) our fine-tuned model outperforms other representative LLMs in generating diverse molecules, including the ones fine-tuned on chemical domains.

Traditional cellular service was designed for global connectivity, but business and logistical constraints led to its fragmentation, with deployments limited to individual countries and regions. Initiatives like Mobile Virtual Network Operators (MVNOs), Mobile Network Aggregators (MNAs), and regulations like ''roam-like-at-home'' have partially restored global service potential, though often at high costs in terms of user bills, application performance, and traffic efficiency. This paper makes two key contributions: first, it surveys the global cellular ecosystem, analyzing the strengths and weaknesses of major players using data from prior research, proprietary datasets, and public sources. Second, it argues that the technology for seamless global service exists in Local Breakout (LBO), a roaming architecture which allows user traffic to be routed directly to the Internet through the visited network, bypassing the home network and/or third-party infrastructures. However, LBO adoption is hindered by issues such as policy enforcement, billing, and Quality of Service (QoS) guarantees, rooted in a lack of trust between operators. The paper concludes by exploring technological advances that could enable LBO, and pave the way for truly global cellular connectivity.

The topological properties of gapped graphene have been explored for valleytronics applications. Prior transport experiments indicated their topological nature through large nonlocal resistance in Hall-bar devices, but the origin of this resistance was unclear. This study focused on dual-gate bilayer graphene (BLG) devices with naturally cleaved edges, examining how edge-etching with an oxygen plasma process affects electron transport. Before etching, local resistance at the charge neutral point increased exponentially with the displacement field and nonlocal resistance was well explained by ohmic contribution, which is typical of gapped BLG. After-etching, however, local resistance saturated with increasing displacement field, and nonlocal resistance deviated by three orders of magnitude from ohmic contribution. We suggest that these significant changes in local and nonlocal resistance arise from the formation of edge conducting pathways after the edge-etching, rather than from a topological property of gapped BLG that has been claimed in previous literatures., Comment: 22 pages, 6 figures

Despite the recent advancements of vision-language-action (VLA) models on a variety of robotics tasks, they suffer from critical issues such as poor generalizability to unseen tasks, due to their reliance on behavior cloning exclusively from successful rollouts. Furthermore, they are typically fine-tuned to replicate demonstrations collected by experts under different settings, thus introducing distribution bias and limiting their adaptability to diverse manipulation objectives, such as efficiency, safety, and task completion. To bridge this gap, we introduce GRAPE: Generalizing Robot Policy via Preference Alignment. Specifically, GRAPE aligns VLAs on a trajectory level and implicitly models reward from both successful and failure trials to boost generalizability to diverse tasks. Moreover, GRAPE breaks down complex manipulation tasks to independent stages and automatically guides preference modeling through customized spatiotemporal constraints with keypoints proposed by a large vision-language model. Notably, these constraints are flexible and can be customized to align the model with varying objectives, such as safety, efficiency, or task success. We evaluate GRAPE across a diverse array of tasks in both real-world and simulated environments. Experimental results demonstrate that GRAPE enhances the performance of state-of-the-art VLA models, increasing success rates on in-domain and unseen manipulation tasks by 51.79% and 60.36%, respectively. Additionally, GRAPE can be aligned with various objectives, such as safety and efficiency, reducing collision rates by 44.31% and rollout step-length by 11.15%, respectively. All code, models, and data are available at https://grape-vla.github.io/, Comment: Website: https://grape-vla.github.io/

The goal of music style transfer is to convert a music performance by one instrument into another while keeping the musical contents unchanged. In this paper, we investigate another style transfer scenario called ``failed-music style transfer''. Unlike the usual music style transfer where the content remains the same and only the instrumental characteristics are changed, this scenario seeks to transfer the music from the source instrument to the target instrument which is deliberately performed off-pitch. Our work attempts to transfer normally played music into off-pitch recorder music, which we call ``failed-style recorder'', and study the results of the conversion. To carry out this work, we have also proposed a dataset of failed-style recorders for this task, called ``FR109 Dataset''. Such an experiment explores the music style transfer task in a more expressive setting, as the generated audio should sound like an ``off-pitch recorder'' while maintaining a certain degree of naturalness., Comment: Accepted by APSIPA 2024

Accurate traversability estimation using an online dense terrain map is crucial for safe navigation in challenging environments like construction and disaster areas. However, traversability estimation for legged robots on rough terrains faces substantial challenges owing to limited terrain information caused by restricted field-of-view, and data occlusion and sparsity. To robustly map traversable regions, we introduce terrain traversability mapping with risk-aware prediction (TRIP). TRIP reconstructs the terrain maps while predicting multi-modal traversability risks, enhancing online autonomous navigation with the following contributions. Firstly, estimating steppability in a spherical projection space allows for addressing data sparsity while accomodating scalable terrain properties. Moreover, the proposed traversability-aware Bayesian generalized kernel (T-BGK)-based inference method enhances terrain completion accuracy and efficiency. Lastly, leveraging the steppability-based Mahalanobis distance contributes to robustness against outliers and dynamic elements, ultimately yielding a static terrain traversability map. As verified in both public and our in-house datasets, our TRIP shows significant performance increases in terms of terrain reconstruction and navigation map. A demo video that demonstrates its feasibility as an integral component within an onboard online autonomous navigation system for quadruped robots is available at https://youtu.be/d7HlqAP4l0c.

We investigate the temperature-dependent deposition of nickelocene (NiCp$_2$) molecules on a single crystal Au(111) substrate, revealing distinct adsorption behaviors and structural formations. At low temperatures (4.2 K), individual NiCp$_2$ molecules adsorb on the herringbone elbows and step edges, forming ordered patterns as molecular coverage increases. However, at 77 K, the molecules dissociate, yielding two main fragments: NiCp fragments that are Ni atoms capped by cyclopentadienyl (Cp) rings, which preferentially adsorb at FCC hollow sites, and Cp radical fragments exhibiting strong substrate interactions. NiCp fragments self-assemble into one-dimensional (1-D) chains along the $\langle 1 1 \bar{2} \rangle$ directions, displaying higher protrusion in STM images. The strain and steric hindrance from the Cp protons induce chiral patterns within the chains, which are well-reproduced by our DFT simulations. In contrast, the Cp fragments maintain distances due to short-range repulsive forces and exhibit low diffusion barriers. Interestingly, the fragments are non-magnetic, as confirmed by both STM measurements and DFT calculations, in contrast to the magnetic signals from intact Nc molecules. In addition to linear chains, dimers of the Ni-Cp fragments form along the $\langle 1 \bar{1} 0\rangle$ directions, requiring gold adatoms for their creation. These results demonstrate the feasibility of constructing complex nanostructures based on metallocenes via on-surface synthesis, opening the possibility for realizing low-dimensional magnetic systems by selecting substrates that preserve the magnetic moment of the fragments., Comment: 8 pages, 9 figures

Leveraging pre-trained models with tailored prompts for in-context learning has proven highly effective in NLP tasks. Building on this success, recent studies have applied a similar approach to the Segment Anything Model (SAM) within a ``one-shot" framework, where only a single reference image and its label are employed. However, these methods face limitations in the medical domain, primarily due to SAM's essential requirement for visual prompts and the over-reliance on pixel similarity for generating them. This dependency may lead to (1) inaccurate prompt generation and (2) clustering of point prompts, resulting in suboptimal outcomes. To address these challenges, we introduce \textbf{Med-PerSAM}, a novel and straightforward one-shot framework designed for the medical domain. Med-PerSAM uses only visual prompt engineering and eliminates the need for additional training of the pretrained SAM or human intervention, owing to our novel automated prompt generation process. By integrating our lightweight warping-based prompt tuning model with SAM, we enable the extraction and iterative refinement of visual prompts, enhancing the performance of the pre-trained SAM. This advancement is particularly meaningful in the medical domain, where creating visual prompts poses notable challenges for individuals lacking medical expertise. Our model outperforms various foundational models and previous SAM-based approaches across diverse 2D medical imaging datasets.

Graphene's exceptional electronic mobility, gate-tunability, and contact transparency with superconducting materials make it ideal for exploring the superconducting proximity effect. However, the work function difference between graphene and superconductors causes unavoidable doping of graphene near contacts, forming a p-n junction in the hole-doped regime and reducing contact transparency. This challenges the device implementation that exploits graphene's bipolarity. To address this limitation, we developed a new fabrication scheme for two-dimensional superconducting contacts that allows independent control over charge concentration and polarity for both the graphene in contact with superconductors and the graphene channel. Contact transparency, conductance enhancement, and Josephson coupling were measured to confirm transparent contacts to both polarities of graphene. Moreover, we demonstrated the Andreev process in the quantum Hall edge state at a negative filling factor of {\nu} = -2. This scheme will open avenues for realizing various theoretical propositions utilizing the bipolarity of graphene combined with superconductivity., Comment: 20 pages, 10 figures

Threat modeling is a crucial component of cybersecurity, particularly for industries such as banking, where the security of financial data is paramount. Traditional threat modeling approaches require expert intervention and manual effort, often leading to inefficiencies and human error. The advent of Large Language Models (LLMs) offers a promising avenue for automating these processes, enhancing both efficiency and efficacy. However, this transition is not straightforward due to three main challenges: (1) the lack of publicly available, domain-specific datasets, (2) the need for tailored models to handle complex banking system architectures, and (3) the requirement for real-time, adaptive mitigation strategies that align with compliance standards like NIST 800-53. In this paper, we introduce ThreatModeling-LLM, a novel and adaptable framework that automates threat modeling for banking systems using LLMs. ThreatModeling-LLM operates in three stages: 1) dataset creation, 2) prompt engineering and 3) model fine-tuning. We first generate a benchmark dataset using Microsoft Threat Modeling Tool (TMT). Then, we apply Chain of Thought (CoT) and Optimization by PROmpting (OPRO) on the pre-trained LLMs to optimize the initial prompt. Lastly, we fine-tune the LLM using Low-Rank Adaptation (LoRA) based on the benchmark dataset and the optimized prompt to improve the threat identification and mitigation generation capabilities of pre-trained LLMs.

Text-based generation and editing of 3D scenes hold significant potential for streamlining content creation through intuitive user interactions. While recent advances leverage 3D Gaussian Splatting (3DGS) for high-fidelity and real-time rendering, existing methods are often specialized and task-focused, lacking a unified framework for both generation and editing. In this paper, we introduce SplatFlow, a comprehensive framework that addresses this gap by enabling direct 3DGS generation and editing. SplatFlow comprises two main components: a multi-view rectified flow (RF) model and a Gaussian Splatting Decoder (GSDecoder). The multi-view RF model operates in latent space, generating multi-view images, depths, and camera poses simultaneously, conditioned on text prompts, thus addressing challenges like diverse scene scales and complex camera trajectories in real-world settings. Then, the GSDecoder efficiently translates these latent outputs into 3DGS representations through a feed-forward 3DGS method. Leveraging training-free inversion and inpainting techniques, SplatFlow enables seamless 3DGS editing and supports a broad range of 3D tasks-including object editing, novel view synthesis, and camera pose estimation-within a unified framework without requiring additional complex pipelines. We validate SplatFlow's capabilities on the MVImgNet and DL3DV-7K datasets, demonstrating its versatility and effectiveness in various 3D generation, editing, and inpainting-based tasks., Comment: Project Page: https://gohyojun15.github.io/SplatFlow/

We demonstrate efficient edge couplers by fabricating a 3D mode size converter on a lithium niobate-on-insulator photonic platform. The 3D mode size converter is fabricated using an etching process that employs a Si external mask to provide height variation and adjust the width variation through tapering patterns via lithography. The measured edge coupling efficiency with a 3D mode size converter was approximately 1.16 dB/facet for the TE mode and approximately 0.71 dB/facet for the TM mode at a wavelength of 1550 nm., Comment: 8 pages, 4 figures

Efficient tokenization of videos remains a challenge in training vision models that can process long videos. One promising direction is to develop a tokenizer that can encode long video clips, as it would enable the tokenizer to leverage the temporal coherence of videos better for tokenization. However, training existing tokenizers on long videos often incurs a huge training cost as they are trained to reconstruct all the frames at once. In this paper, we introduce CoordTok, a video tokenizer that learns a mapping from coordinate-based representations to the corresponding patches of input videos, inspired by recent advances in 3D generative models. In particular, CoordTok encodes a video into factorized triplane representations and reconstructs patches that correspond to randomly sampled $(x,y,t)$ coordinates. This allows for training large tokenizer models directly on long videos without requiring excessive training resources. Our experiments show that CoordTok can drastically reduce the number of tokens for encoding long video clips. For instance, CoordTok can encode a 128-frame video with 128$\times$128 resolution into 1280 tokens, while baselines need 6144 or 8192 tokens to achieve similar reconstruction quality. We further show that this efficient video tokenization enables memory-efficient training of a diffusion transformer that can generate 128 frames at once., Comment: Code is available on the project webpage: https://huiwon-jang.github.io/coordtok/

We report measurements of the absolute branching fractions $\mathcal{B}(B_s^0 \to D_s^{\pm} X)$, $\mathcal{B}(B_s^0 \to D^0/\bar{D}^0 X)$, and $\mathcal{B}(B_s^0 \to D^{\pm} X)$, where the latter is measured for the first time. The results are based on a 121.4\,fb$^{-1}$ data sample collected at the $\Upsilon(10860)$ resonance by the Belle detector at the KEKB asymmetric-energy $e^+ e^-$ collider. We reconstruct one $B_s^0$ meson in $e^+e^- \to \Upsilon(10860) \to B_s^{*} \bar{B}_s^{*}$ events and measure yields of $D_s^+$, $D^0$, and $D^+$ mesons in the rest of the event. We obtain $\mathcal{B}(B_s^0 \to D_s^{\pm} X) = (68.6 \pm 7.2 \pm 4.0)\%$, $\mathcal{B}(B_s^0 \to D^0/\bar{D}^0 X) = (21.5 \pm 6.1 \pm 1.8)\%$, and $\mathcal{B}(B_s^0 \to D^{\pm} X) = (12.6 \pm 4.6 \pm 1.3)\%$, where the first uncertainty is statistical and the second is systematic. Averaging with previous Belle measurements gives $\mathcal{B}(B_s^0 \to D_s^{\pm} X) = (63.4 \pm 4.5 \pm 2.2)\%$ and $\mathcal{B}(B_s^0 \to D^0/\bar{D}^0 X) = (23.9 \pm 4.1 \pm 1.8)\%$. For the $B_s^0$ production fraction at the $\Upsilon(10860)$, we find $f_s = (21.4^{+1.5}_{-1.7})\%$., Comment: 23 pages, 9 figures, submitted to JHEP

When malware employs an unseen zero-day exploit, traditional security measures such as vulnerability scanners and antivirus software can fail to detect them. This is because these tools rely on known patches and signatures, which do not exist for new zero-day attacks. Furthermore, existing machine learning methods, which are trained on specific and occasionally outdated malware samples, may struggle to adapt to features in new malware. To address this issue, there is a need for a more robust machine learning model that can identify relationships between malware samples without being trained on a particular malware feature set. This is particularly crucial in the field of cybersecurity, where the number of malware samples is limited and obfuscation techniques are widely used. Current approaches using stacked autoencoders aim to remove the noise introduced by obfuscation techniques through reconstruction of the input. However, this approach ignores the semantic relationships between features across different malware samples. To overcome this limitation, we propose a novel Siamese Neural Network (SNN) that uses relation-aware embeddings to calculate more accurate similarity probabilities based on semantic details of different malware samples. In addition, by using entropy images as inputs, our model can extract better structural information and subtle differences in malware signatures, even in the presence of obfuscation techniques. Evaluations on two large malware sample sets using the N-shot and N-way methods show that our proposed model is highly effective in predicting previously unseen malware, even in the presence of obfuscation techniques.

Test-time adaptation (TTA) is an effective approach to mitigate performance degradation of trained models when encountering input distribution shifts at test time. However, existing TTA methods often suffer significant performance drops when facing additional class distribution shifts. We first analyze TTA methods under label distribution shifts and identify the presence of class-wise confusion patterns commonly observed across different covariate shifts. Based on this observation, we introduce label Distribution shift-Aware prediction Refinement for Test-time adaptation (DART), a novel TTA method that refines the predictions by focusing on class-wise confusion patterns. DART trains a prediction refinement module during an intermediate time by exposing it to several batches with diverse class distributions using the training dataset. This module is then used during test time to detect and correct class distribution shifts, significantly improving pseudo-label accuracy for test data. Our method exhibits 5-18% gains in accuracy under label distribution shifts on CIFAR-10C, without any performance degradation when there is no label distribution shift. Extensive experiments on CIFAR, PACS, OfficeHome, and ImageNet benchmarks demonstrate DART's ability to correct inaccurate predictions caused by test-time distribution shifts. This improvement leads to enhanced performance in existing TTA methods, making DART a valuable plug-in tool.

We theoretically and numerically elucidate the electrical control over spin waves in antiferromagnetic materials (AFM) with biaxial anisotropies and Dzyaloshinskii-Moriya interactions. The spin wave dispersion in an AFM manifests as a bifurcated spectrum with distinct high-frequency and low-frequency bands. Utilizing a heterostructure comprised of platinum and the AFM, we demonstrate anisotropic control of spin-wave bands via spin currents with three-dimensional spin polarizations, encompassing both resonant and propagating wave modes. Moreover, leveraging the confined geometry, we explore the possibility of controlling spin waves within a spectral domain ranging from tens of gigahertz to sub-terahertz frequencies. The implications of our findings suggest the potential for developing a terahertz wave source with electrical tunability, thereby facilitating its incorporation into ultrafast, broadband, and wireless communication technologies.

In 1995, Ismail and Masson introduced orthogonal polynomials of types \( R_I \) and \( R_{II} \), which are defined by specific three-term recurrence relations with additional conditions. Recently, Kim and Stanton found a combinatorial interpretation for the moments of orthogonal polynomials of type \( R_I \) in the spirit of the combinatorial theory of orthogonal polynomials due to Flajolet and Viennot. In this paper, we push this combinatorial model further to orthogonal polynomials of type \( R_{II} \). Moreover, we generalize orthogonal polynomials of type \( R_{II} \) by relaxing some of their conditions. We then prove a master theorem, which generalizes combinatorial models for moments of various types of orthogonal polynomials: classical orthogonal polynomials, Laurent biorthogonal polynomials, and orthogonal polynomials of types \( R_I \) and \( R_{II} \).

We experimentally demonstrate magnetic steganography using wide field quantum microscopy based on diamond nitrogen vacancy centers. The method offers magnetic imaging capable of revealing concealed information otherwise invisible with conventional optical measurements. For a proof of principle demonstration of the magnetic steganography, micrometer structures designed as pixel arts, barcodes, and QR codes are fabricated using mixtures of magnetic and nonmagnetic materials, nickel and gold. We compare three different imaging modes based on the changes in frequency, linewidth, and contrast of the NV electron spin resonance, and find that the last mode offers the best quality of reconstructing hidden magnetic images. By simultaneous driving of the NV qutrit states with two independent microwave fields, we expediate the imaging time by a factor of three. This work shows potential applications of quantum magnetic imaging in the field of image steganography., Comment: 26 pages, 6 figures

Estimating the homography between two images is crucial for mid- or high-level vision tasks, such as image stitching and fusion. However, using supervised learning methods is often challenging or costly due to the difficulty of collecting ground-truth data. In response, unsupervised learning approaches have emerged. Most early methods, though, assume that the given image pairs are from the same camera or have minor lighting differences. Consequently, while these methods perform effectively under such conditions, they generally fail when input image pairs come from different domains, referred to as multimodal image pairs. To address these limitations, we propose AltO, an unsupervised learning framework for estimating homography in multimodal image pairs. Our method employs a two-phase alternating optimization framework, similar to Expectation-Maximization (EM), where one phase reduces the geometry gap and the other addresses the modality gap. To handle these gaps, we use Barlow Twins loss for the modality gap and propose an extended version, Geometry Barlow Twins, for the geometry gap. As a result, we demonstrate that our method, AltO, can be trained on multimodal datasets without any ground-truth data. It not only outperforms other unsupervised methods but is also compatible with various architectures of homography estimators. The source code can be found at:~\url{https://github.com/songsang7/AltO}, Comment: This paper is accepted to the Thirty-Eighth Annual Conference on Neural Information Processing Systems (NeurIPS 2024)

Predicting and constructing road geometric information (e.g., lane lines, road markers) is a crucial task for safe autonomous driving, while such static map elements can be repeatedly occluded by various dynamic objects on the road. Recent studies have shown significantly improved vectorized high-definition (HD) map construction performance, but there has been insufficient investigation of temporal information across adjacent input frames (i.e., clips), which may lead to inconsistent and suboptimal prediction results. To tackle this, we introduce a novel paradigm of clip-level vectorized HD map construction, MapUnveiler, which explicitly unveils the occluded map elements within a clip input by relating dense image representations with efficient clip tokens. Additionally, MapUnveiler associates inter-clip information through clip token propagation, effectively utilizing long-term temporal map information. MapUnveiler runs efficiently with the proposed clip-level pipeline by avoiding redundant computation with temporal stride while building a global map relationship. Our extensive experiments demonstrate that MapUnveiler achieves state-of-the-art performance on both the nuScenes and Argoverse2 benchmark datasets. We also showcase that MapUnveiler significantly outperforms state-of-the-art approaches in a challenging setting, achieving +10.7% mAP improvement in heavily occluded driving road scenes. The project page can be found at https://mapunveiler.github.io., Comment: 18 pages, 9 figures, NeurIPS 2024

In recent years, with the rapid development of augmented reality (AR) technology, there is an increasing demand for multi-user collaborative experiences. Unlike for single-user experiences, ensuring the spatial localization of every user and maintaining synchronization and consistency of positioning and orientation across multiple users is a significant challenge. In this paper, we propose a multi-user localization system based on ORB-SLAM2 using monocular RGB images as a development platform based on the Unity 3D game engine. This system not only performs user localization but also places a common virtual object on a planar surface (such as table) in the environment so that every user holds a proper perspective view of the object. These generated virtual objects serve as reference points for multi-user position synchronization. The positioning information is passed among every user's AR devices via a central server, based on which the relative position and movement of other users in the space of a specific user are presented via virtual avatars all with respect to these virtual objects. In addition, we use deep learning techniques to estimate the depth map of an image from a single RGB image to solve occlusion problems in AR applications, making virtual objects appear more natural in AR scenes.

Sufficient overlap of propensity scores is one of the most critical assumptions in observational studies. In this article, we will cover the severity in statistical inference under such assumption failure with weighting, one of the most dominating causal inference methodologies. Then we propose a simple, yet novel remedy: "mixing" the treated and control groups in the observed dataset. We state that our strategy has three key advantages: (1) Improvement in estimators' accuracy especially in weak overlap, (2) Identical targeting population of treatment effect, (3) High flexibility. We introduce a property of mixed sample that offers a safer inference by implementing onto both traditional and modern weighting methods. We illustrate this with several extensive simulation studies and guide the readers with a real-data analysis for practice., Comment: 30 pages, 6 figures

Atom probe tomography (APT) enables near atomic scale three dimensional elemental mapping through the controlled field evaporation of surface atoms triggered by the combined application of a DC voltage and either voltage or laser pulses. As the selected laser wavelength for the atom probes transitioned from the near-infrared (1050 nm) to shorter wavelengths e.g., green (532 nm) and near ultraviolet, the quality of data improved and the range of materials amenable for analysis broadened. A new commercial laser atom probe with a wavelength of 257.5 nm, referred to as deep ultraviolet (DUV), has been recently launched. However, the effects of DUV lasers on different classes of materials have not yet been systematically investigated. In this study, a range of materials, including metals, semiconductor, and oxides, have been examined using commercial atom probes with different laser wavelengths but in principle comparable particle detection system. The quality of the NUV and DUV laser atom probe data is evaluated based on four key metrics: background, detection events, ion detection histogram, and mass-resolving power. Furthermore, the application of a thin coating to the finished APT specimens enhances the data quality for both laser wavelengths.

We present measurements of $B \to K{}^{*}(892)\gamma$ decays using $365\,{\rm fb}^{-1}$ of data collected from 2019 to 2022 by the Belle~II experiment at the SuperKEKB asymmetric-energy $e^+e^-$ collider. The data sample contains $(387 \pm 6) \times 10^6$ $B\overline{B}$ events. We measure branching fractions ($\mathcal{B}$) and $C\!P$ asymmetries ($\mathcal{A}_{C\!P}$) for both $B^{0}\to K{}^{*0}\gamma$ and $B^{+}\to K{}^{*+}\gamma$ decays. The difference in $C\!P$ asymmetries ($\Delta \mathcal{A}_{C\!P}$) and the isospin asymmetry ($\Delta_{0+}$) between these neutral and charged channels are also measured. We obtain the following branching fractions and $C\!P$ asymmetries: $\mathcal{B} (B^{0} \to K{}^{*0}\gamma) = (4.14 \pm 0.10 \pm 0.11 ) \times 10^{-5}$, $\mathcal{B} (B^{+} \to K{}^{*+}\gamma) = (4.02 \pm 0.13 \pm 0.13 )\times 10^{-5}$, $\mathcal{A}_{C\!P} (B^{0} \to K{}^{*0}\gamma) = (-3.3 \pm 2.3 \pm 0.4 )\%$, and $\mathcal{A}_{C\!P} (B^{+} \to K{}^{*+}\gamma) = (-0.7 \pm 2.9 \pm 0.6 )\%$. The measured difference in $C\!P$ asymmetries is $\Delta \mathcal{A}_{C\!P} = (+2.6 \pm 3.8 \pm 0.7 )\%$, and the measured isospin asymmetry is $\Delta_{0+} = (+5.0 \pm 2.0 \pm 1.5 )\%$. The first uncertainties listed are statistical and the second are systematic. These results are consistent with world-average values and theory predictions.

Multi-view learning often faces challenges in effectively leveraging images captured from different angles and locations. This challenge is particularly pronounced when addressing inconsistencies and uncertainties between views. In this paper, we propose a novel Multi-View Uncertainty-Weighted Mutual Distillation (MV-UWMD) method. Our method enhances prediction consistency by performing hierarchical mutual distillation across all possible view combinations, including single-view, partial multi-view, and full multi-view predictions. This introduces an uncertainty-based weighting mechanism through mutual distillation, allowing effective exploitation of unique information from each view while mitigating the impact of uncertain predictions. We extend a CNN-Transformer hybrid architecture to facilitate robust feature learning and integration across multiple view combinations. We conducted extensive experiments using a large, unstructured dataset captured from diverse, non-fixed viewpoints. The results demonstrate that MV-UWMD improves prediction accuracy and consistency compared to existing multi-view learning approaches.

We introduce DAHL, a benchmark dataset and automated evaluation system designed to assess hallucination in long-form text generation, specifically within the biomedical domain. Our benchmark dataset, meticulously curated from biomedical research papers, consists of 8,573 questions across 29 categories. DAHL evaluates fact-conflicting hallucinations in Large Language Models (LLMs) by deconstructing responses into atomic units, each representing a single piece of information. The accuracy of these responses is averaged to produce the DAHL Score, offering a more in-depth evaluation of hallucinations compared to previous methods that rely on multiple-choice tasks. We conduct experiments with 8 different models, finding that larger models tend to hallucinate less; however, beyond a model size of 7 to 8 billion parameters, further scaling does not significantly improve factual accuracy. The DAHL Score holds potential as an efficient alternative to human-annotated preference labels, being able to be expanded to other specialized domains. We release the dataset and code in public., Comment: EMNLP2024/FEVER

The remarkable advances in deep learning have led to the emergence of many off-the-shelf classifiers, e.g., large pre-trained models. However, since they are typically trained on clean data, they remain vulnerable to adversarial attacks. Despite this vulnerability, their superior performance and transferability make off-the-shelf classifiers still valuable in practice, demanding further work to provide adversarial robustness for them in a post-hoc manner. A recently proposed method, denoised smoothing, leverages a denoiser model in front of the classifier to obtain provable robustness without additional training. However, the denoiser often creates hallucination, i.e., images that have lost the semantics of their originally assigned class, leading to a drop in robustness. Furthermore, its noise-and-denoise procedure introduces a significant distribution shift from the original distribution, causing the denoised smoothing framework to achieve sub-optimal robustness. In this paper, we introduce Fine-Tuning with Confidence-Aware Denoised Image Selection (FT-CADIS), a novel fine-tuning scheme to enhance the certified robustness of off-the-shelf classifiers. FT-CADIS is inspired by the observation that the confidence of off-the-shelf classifiers can effectively identify hallucinated images during denoised smoothing. Based on this, we develop a confidence-aware training objective to handle such hallucinated images and improve the stability of fine-tuning from denoised images. In this way, the classifier can be fine-tuned using only images that are beneficial for adversarial robustness. We also find that such a fine-tuning can be done by updating a small fraction of parameters of the classifier. Extensive experiments demonstrate that FT-CADIS has established the state-of-the-art certified robustness among denoised smoothing methods across all $\ell_2$-adversary radius in various benchmarks., Comment: 26 pages; TMLR 2024; Code is available at https://github.com/suhyeok24/FT-CADIS

Semiconductor heterojunctions are foundational to many advanced electronic and optoelectronic devices. However, achieving high-quality, lattice-mismatched interfaces remains challenging, limiting both scalability and device performance. Semiconductor grafting offers a promising solution by directly forming electrically active, lattice-mismatched heterojunctions between dissimilar materials. However, its scalability and uniformity at the wafer level have yet to be demonstrated. This work demonstrates the achievement of highly uniform, reproducible results across silicon, sapphire, and gallium nitride (GaN) substrates using wafer-scale semiconductor grafting. To illustrate this scalability, we conducted an in-depth study of a grafted Si/GaN heterojunction, examining band alignment through X-ray photoelectron spectroscopy and confirming crystallinity and interfacial integrity with scanning transmission electron microscopy. The resulting p-n diodes exhibit significantly enhanced electrical performance and wafer-scale uniformity compared to conventional approaches. This work establishes wafer-scale semiconductor grafting as a versatile and scalable technology, bridging the gap between laboratory-scale research and industrial manufacturing for heterogeneous semiconductor integration, and paving the way for novel, high-performance electronic and optoelectronic devices., Comment: 23 pages, 6 figures

We present Reusable Motion prior (ReMP), an effective motion prior that can accurately track the temporal evolution of motion in various downstream tasks. Inspired by the success of foundation models, we argue that a robust spatio-temporal motion prior can encapsulate underlying 3D dynamics applicable to various sensor modalities. We learn the rich motion prior from a sequence of complete parametric models of posed human body shape. Our prior can easily estimate poses in missing frames or noisy measurements despite significant occlusion by employing a temporal attention mechanism. More interestingly, our prior can guide the system with incomplete and challenging input measurements to quickly extract critical information to estimate the sequence of poses, significantly improving the training efficiency for mesh sequence recovery. ReMP consistently outperforms the baseline method on diverse and practical 3D motion data, including depth point clouds, LiDAR scans, and IMU sensor data. Project page is available in https://hojunjang17.github.io/ReMP., Comment: 8 main pages, WACV 2025

In Special Relativity, massless objects are characterized as either vacuum states or as radiation propagating at the speed of light. This distinction extends to General Relativity for asymptotically flat initial data sets (IDS) \((M^n, g, k)\), where vacuum is represented by slices of Minkowski space, and radiation is modeled by slices of \(pp\)-wave spacetimes. In contrast, we demonstrate that asymptotically hyperboloidal IDS with zero mass must embed isometrically into Minkowski space, with no possible IDS configurations modeling radiation in this setting. Our result holds under the most general assumptions. The proof relies on precise decay estimates for spinors on level sets of spacetime harmonic functions and works in all dimensions., Comment: 23 pages; comments welcome!

Fine-tuning pre-trained models for downstream tasks is a widely adopted technique known for its adaptability and reliability across various domains. Despite its conceptual simplicity, fine-tuning entails several troublesome engineering choices, such as selecting hyperparameters and determining checkpoints from an optimization trajectory. To tackle the difficulty of choosing the best model, one effective solution is model fusion, which combines multiple models in a parameter space. However, we observe a large discrepancy between loss and metric landscapes during the fine-tuning of pre-trained language models. Building on this observation, we introduce a novel model fusion technique that optimizes both the desired metric and loss through multi-objective Bayesian optimization. In addition, to effectively select hyperparameters, we establish a two-stage procedure by integrating Bayesian optimization processes into our framework. Experiments across various downstream tasks show considerable performance improvements using our Bayesian optimization-guided method.

Acoustic detection has many applications across science and technology, from medical to imaging and communications. However, most acoustic sensors have a common limitation in that the detection must be near the acoustic source. Alternatively laser interferometry with picometer-scale motional displacement detection can rapidly and precisely measure sound induced minute vibrations on remote surfaces. Here we demonstrate the feasibility of sound detection up to 100 kHz at remote sites with ~ 60 m of optical path length via laser homodyne interferometry. Based on our ultrastable Hz-linewidth laser with 10-15 fractional stability, our laser interferometer achieves 0.5 pm/Hz1/2 displacement sensitivity near 10 kHz, bounded only by laser frequency noise over 10 kHz. Between 140 Hz to 15 kHz, we achieve a homodyne acoustic sensing sensitivity of sub-nm/Pa across our conversational frequency overtones. The minimal sound pressure detectable over 60 m of optical path length is ~ 2 mPa, with dynamic ranges over 100 dB. With the demonstrated standoff picometric distance metrology, we successfully detected and reconstructed musical scores of normal conversational volumes with high fidelity. The acoustic detection via this precision laser interferometer could be applied to selective area sound sensing for remote acoustic metrology, optomechanical vibrational motion sensing and ultrasensitive optical microphones at the laser frequency noise limits., Comment: 25 pages, 10 figures

The formation pathways of lenticular galaxies (S0s) in field environments remain a matter of debate. We utilize the cosmological hydrodynamic simulation, NewHorizon, to investigate the issue. We select two massive star-formation quenched S0s as our main sample. By closely tracing their physical and morphological evolution, we identify two primary formation channels: mergers and counter-rotating gas accretion. The former induces central gas inflow due to gravitational and hydrodynamic torques, triggering active central star formation which quickly depletes the gas of the galaxy. Counter-rotating gas accretion overall has a similar outcome but more exclusively through hydrodynamic collisions between the pre-existing and newly-accreted gas. Both channels lead to S0 morphology, with gas angular momentum cancellation being a crucial mechanism. These formation pathways quench star formation on a short timescale (< Gyr) compared to the timescales of environmental effects. We also discuss how counter-rotating gas accretion may explain the origin of S0s with ongoing star formation and the frequently observed gas-star misaligned kinematics in S0s., Comment: 19 pages, 16 figures, Accepted for publication in ApJ, Oct 27 2024

In this paper, we give a characterization of Fano type varieties of plc pairs. This extends a result due to Choi and Park for pklt pairs. Moreover, we show that for a plc pair $(X,\Delta)$, if no plc centers are contained in the augmented base locus $\mathbf{B}_{+}(-(K_X+\Delta))$, then $(X,\Delta)$ admits a good $-(K_X+\Delta)$-minimal model.

For a circle action on a compact almost complex manifold with a fixed point, the lower bound on the number of fixed points is known in dimension up to 12 except 10. In this paper, we show that if the circle group acts on a 10-dimensional compact almost complex manifold with a fixed point, then there are at least 6 fixed points. This minimum is attained by $\mathbb{CP}^5$ and $S^6 \times \mathbb{CP}^2$. We establish this lower bound by showing that there does not exist a circle action on a 10-dimensional compact almost complex manifold with 4 fixed points.

We present a search for long-lived particles (LLPs) produced from kaon decay that decay to two muons inside the ICARUS neutrino detector. This channel would be a signal of hidden sector models that can address outstanding issues in particle physics such as the strong CP problem and the microphysical origin of dark matter. The search is performed with data collected in the Neutrinos at the Main Injector (NuMI) beam at Fermilab corresponding to $2.41\times 10^{20}$ protons-on-target. No new physics signal is observed, and we set world-leading limits on heavy QCD axions, as well as for the Higgs portal scalar among dedicated searches. Limits are also presented in a model-independent way applicable to any new physics model predicting the process $K\to \pi+S(\to\mu\mu)$, for a long-lived particle S. This result is the first search for new physics performed with the ICARUS detector at Fermilab. It paves the way for the future program of long-lived particle searches at ICARUS.

Metallic glass is a frozen liquid with structural disorder that retains degenerate free energy without spontaneous symmetry breaking to become a solid. For over half a century, this puzzling structure has raised fundamental questions about how structural disorder impacts glass-liquid phase transition kinetics, which remain elusive without direct evidence. In this study, through single-pulse, time-resolved imaging using X-ray free-electron lasers, we visualized the glass-to-liquid transition, revealing a previously hidden dormant phase that does not involve any macroscopic volume change within the crossover regime between the two phases. Although macroscopically inactive, nanoscale redistribution occurs, forming channeld low-density bands within this dormant phase that drives the glass transition. By providing direct microscopic evidence, this work presents a new perspective on the phase transition process in disordered materials, which can be extended to various liquid and solid phases in other complex systems., Comment: 25 pages, 4 figures

To meet the growing demand for accurate, rapid, and cost-effective at-home clinical testing, we developed a radical-mediated enzyme assay (REEA) integrated with a paper fluidic system and electrically read by a handheld field-effect transistor (FET) device. The REEA utilizes horseradish peroxidase (HRP) to catalyze the conversion of aromatic substrates into radical forms, producing protons detected by an ion-sensitive FET for biomarker quantification. Through screening 14 phenolic compounds, halogenated phenols emerged as optimal substrates for the REEA. Encased in an affordable cartridge ($0.55 per test), the system achieved a detection limit of 146 fg/mL for estradiol (E2), with a coefficient of variation (CV) below 9.2% in E2-spiked samples and an r2 of 0.963 across a measuring range of 19 to 4,551 pg/mL in clinical plasma samples, providing results in under 10 minutes. This adaptable system not only promises to offer a fast and reliable platform, but also holds significant potential for expansion to a wide array of biomarkers, paving the way for broader clinical and home-based applications.

The rapid advancement of Large Language Models (LLMs) necessitates robust evaluation methodologies. Current benchmarking approaches often rely on comparing model outputs against predefined prompts and reference outputs. Relying on predefined reference outputs hinders flexible adaptation of benchmarks to the rapidly evolving capabilities of LLMs. This limitation necessitates periodic efforts to prepare new benchmarks. To keep pace with rapidly evolving LLM capabilities, we propose a more flexible benchmarking approach. Our method, \textit{\textbf{Varco Arena}}, provides reference-free benchmarking of LLMs in tournament style. \textit{\textbf{Varco Arena}} directly compares LLM outputs across a diverse set of prompts, determining model rankings through a single-elimination tournament structure. This direct pairwise comparison offers two key advantages: (1) Direct comparison, unmediated by reference text, more effectively orders competing LLMs, resulting in more reliable rankings, and (2) reference-free approach to benchmarking adds flexibility in updating benchmark prompts by eliminating the need for quality references. Our empirical results, supported by simulation experiments, demonstrate that the \textit{\textbf{Varco Arena}} tournament approach aligns better with the current Elo model for benchmarking LLMs. The alignment is measured in terms of Spearman correlation, showing improvement over current practice of benchmarking that use reference outputs as comparison \textit{anchor}s., Comment: 7 pages for main body, 13 pages in total

The conventional targeted adversarial attacks add a small perturbation to an image to make neural network models estimate the image as a predefined target class, even if it is not the correct target class. Recently, for visual-language models (VLMs), the focus of targeted adversarial attacks is to generate a perturbation that makes VLMs answer intended target text outputs. For example, they aim to make a small perturbation on an image to make VLMs' answers change from "there is an apple" to "there is a baseball." However, answering just intended text outputs is insufficient for tricky questions like "if there is a baseball, tell me what is below it." This is because the target of the adversarial attacks does not consider the overall integrity of the original image, thereby leading to a lack of visual reasoning. In this work, we focus on generating targeted adversarial examples with visual reasoning against VLMs. To this end, we propose 1) a novel adversarial attack procedure -- namely, Replace-then-Perturb and 2) a contrastive learning-based adversarial loss -- namely, Contrastive-Adv. In Replace-then-Perturb, we first leverage a text-guided segmentation model to find the target object in the image. Then, we get rid of the target object and inpaint the empty space with the desired prompt. By doing this, we can generate a target image corresponding to the desired prompt, while maintaining the overall integrity of the original image. Furthermore, in Contrastive-Adv, we design a novel loss function to obtain better adversarial examples. Our extensive benchmark results demonstrate that Replace-then-Perturb and Contrastive-Adv outperform the baseline adversarial attack algorithms. We note that the source code to reproduce the results will be available., Comment: 13 pages, 5 figure

We describe Elevator, a unifying polymorphic foundation for metaprogramming with memory management based on adjoint modalities. In this setting, we distinguish between multiple memory regions using modes where each mode has a specific set of structural properties. This allows us not only to capture linear (i.e. garbage-free) memory regions and (ordinary) intuitionistic (i.e. garbage-collected or persistent) memory regions, but also to capture accessibility between the memory regions using a preorder between modes. This preorder gives us the power to describe monadic and comonadic programming. As a consequence, it extends the existing logical view of metaprogramming in two directions: first, it ensures that code generation can be done efficiently by controlling memory accesses; second, it allows us to provide resource guarantees about the generated code (i.e. code that is for example garbage-free). We present the static and dynamic semantics of Elevator. In particular, we prove the substructurality of variable references and type safety of the language. We also establish mode safety, which guarantees that the evaluation of a term does not access a value in an inaccessible memory.

We measure the time-integrated CP asymmetry in $D^{0} \rightarrow K^{0}_{S}K^{0}_{S}$ decays reconstructed in $e^{+}e^{-} \rightarrow c\overline{c}$ events collected by the Belle and Belle II experiments. The corresponding data samples have integrated luminosities of 980 fb$^{-1}$ and 428 fb$^{-1}$, respectively. The $D^{0}$ decays are required to originate from the $D^{*+} \rightarrow D^{0}\pi^{+}$ decay, which determines the charm flavor at production time. A control sample of $D^{0} \rightarrow K^{+}K^{-}$ decays is used to correct for production and detection asymmetries. The result, $(-1.4\pm1.3{\rm(stat)}\pm0.1{\rm (syst)})\%$, is consistent with previous determinations and with CP symmetry., Comment: 10 pages, 3 figures. arXiv admin note: text overlap with arXiv:2410.22961