An adaptive spatial approximation algorithm is proposed while minimizing false contouring based on local image characteristics considering human visual systems to save the amount of image frame memory and power for image data transmission. The proposed algorithm is generalized by using block-based k-means clustering to categorize image blocks with the same characteristics, and the amount of approximation is evaluated for each cluster. Two different image quality standards are maintained to maximize image approximation while maintaining the required image qualities. The proposed algorithm can reduce frame buffer memory up to 35.11% on average.

A family of sets is a collection where each element is a set, enabling the representation of many practical concepts. Various operations on families of sets are widely applied in fields such as databases and data mining. Since the size of set families in these applications often becomes exponentially large, we need sophisticated algorithms to manipulate them. Zero-suppressed decision diagrams (ZDDs) efficiently represent families of sets using directed acyclic graphs, supporting various operations known as family algebra. However, designing efficient algorithms for ZDDs demands expertise and is costly, underscoring the need for more accessible design methods.

This paper introduces an algorithm template that extends ZDD-based family algebra. We can easily design new operations by setting component functions to the template. The template is a natural generalization of existing operations, reproducing them without loss of efficiency. Additionally, it enables the generation of previously impractical ZDD operations without deep knowledge of ZDDs. This paper also presents concrete examples of new operations.

As a downstream task of visual entity and relationship extraction, human-object interaction detection focuses on complex relationships centered around humans as the primary subject. This has significant potential for application in some labour-intensive industries such as construction engineering. However, the data in these contexts often display a long-tailed distribution, featuring numerous unknown entities and relationships that are not present in standard datasets. This phenomenon places considerable demands on the model's zero-shot learning capabilities. To tackle this challenge, this letter proposed an end-to-end human-object interaction detection method that utilized domain knowledge graph embeddings as part of prior queries for the decoders. In the case study, this method achieved a mean Average Precision (mAP) of 48.57% for the Full types across various scenarios. Specifically, the Rare types achieved a mAP of 52.45%, while the Non-Rare types achieved a mAP of 41.67%.

As one of the most widely used programming languages in modern software development, Python hosts a vast open-source codebase on GitHub, where code reuse is widespread. This study leverages open-source Python projects from GitHub and applies automated testing to discover pairs of functionally equivalent methods. We collected and processed methods from 5,100 Python repositories, but Python's lack of static type checking presented unique challenges for grouping these methods. To address this, we conducted detailed type inference and organized methods based on their inferred types, providing a structured foundation for subsequent analysis. We then employed automated test generation to produce unit tests for each method, running them against one another within their respective groups to identify candidate pairs that yielded identical outputs from the same inputs. Through manual verification, we ultimately identified 68 functionally equivalent method pairs and 683 functionally non-equivalent pairs. These pairs were compiled into a comprehensive dataset, serving as the basis for further examination. With this dataset, we not only evaluated the ability of large language models (LLMs) to recognize functional equivalence, evaluating both their accuracy and the challenges posed by diverse implementations, but also conducted a systematic performance evaluation of equivalent methods, measuring execution times and analyzing the underlying causes of efficiency differences. The findings demonstrate the potential of LLMs to identify functionally equivalent methods and highlight areas requiring further advancement.

Multi-objective reinforcement learning (MORL) addresses sequential decision-making problems involving conflicting objectives. While most existing methods assume access to known or explicitly defined utility functions, many real-world tasks feature implicit, nonlinear utilities that are only available as delayed black-box feedback. To tackle this challenge, we propose Adaptive Multi-Objective Actor-Critic (AMOAC), a scalable framework that dynamically aligns policy optimization with implicit utility signals, without requiring prior knowledge of the utility function's form. AMOAC employs a multi-critic architecture to maintain computational efficiency as the number of objectives grows, and introduces a dynamic direction-aligned weighting mechanism to guide policy updates toward utility maximization. Experiments on benchmark environments—including Deep Sea Treasure, Minecart, and Four Room—demonstrate that AMOAC consistently matches or exceeds the performance of baselines with explicit utility access, achieving robust adaptation and convergence under both linear and nonlinear utility scenarios. These results highlight the potential of dynamic weight adjustment in MORL for handling implicit preference structures and limited feedback settings.

The proliferation of fake accounts in social networks has prompted growing attention to the development of effective detection techniques for ensuring cyberspace security. These fake accounts frequently employ sophisticated camouflage strategies to evade detection, which compromises the reliability of local neighborhood information. We propose GRFA, a novel approach for fake account detection that incorporates similarity-based adaptive graph reconstruction. The framework introduces a reinforcement learning-based adaptive mechanism to construct similarity edges, which dynamically refines the graph structure to better capture global dependencies. These refined structures are then incorporated into a heterogeneous graph neural network with dual aggregation, significantly improving the detection of camouflaged accounts. Experimental results demonstrate that GRFA outperforms state-of-the-art methods across multiple real-world datasets.

Recently, websites that enable users to share and search for cooking recipes have gained popularity. Each recipe typically includes various pieces of information, including a title, a list of ingredients, and detailed steps described in text and illustrated with photos. The estimated cooking time for each recipe is another valuable information when selecting a recipe. However, it can be difficult to accurately determine cooking time because it depends on various factors, such as heat level, ingredient quantity, and cooking skill level. Therefore, some recipes do not include information on cooking time. In this study, we consider the prediction of cooking time in general scenarios based on a list of ingredients and a textual description of each recipe's cooking process using BERT, a natural language processing model. To this end, we propose an additional pretraining method that assigns greater weight to words related to cooking time using a cooking ontology. Our experimental results show that our method outperforms a fine-tuned BERT model with additional pre-training using a commonly employed approach. Notably, words representing “Kitchen Tools” are particularly associated with cooking time.

Ellipse detection plays a critical role in fields such as medical diagnosis, environmental monitoring, and industrial automation. However, traditional methods (e.g., Hough transform, least-squares fitting, and edge-following techniques) suffer from high computational complexity and poor noise robustness. To address these limitations, we propose a hybrid framework that integrates deep learning with geometric constraints. First, Faster RCNN is employed to localize axis-aligned bounding boxes (AABBs) of ellipses. Then, a point-pair filtering strategy extracts edge points satisfying predefined geometric constraints, followed by weighted least-squares fitting to estimate ellipse parameters. Compared with traditional approaches, our method directly identifies AABBs, significantly enhancing both the efficiency and accuracy of multi-target ellipse detection in practice. Experiments are conducted on two synthetic datasets. The results show that our proposed method achieves superior precision and F-measure compared to conventional ellipse detection algorithms.

Developers rely on third-party open-source libraries to save time and reuse well-tested code. As technology stacks diversify, libraries are deployed across multiple ecosystems to reach broader audiences and accommodate different user needs. However, maintainers may hesitate due to concerns about increased maintenance effort and uncertain adoption outcomes. This study investigates the impact of cross-ecosystem deployments on maintenance effort and project adoption. Analyzing 972,592 NPM and PyPI packages, we focused on 420 actively maintained libraries that exist in both ecosystems. Of these, 184 were initially deployed to NPM, 148 to PyPI, and 88 were synchronized releases. We collected GitHub metrics—including issues, pull requests, contributors, forks, and commits—over a three-month period before and after deployment. Results show that 80-85% of packages saw no major maintenance activity. However, synchronized releases led to a 15.91% rise in issue reporting and an 11.49% increase in pull requests (PyPI → NPM), indicating higher initial maintenance effort. Popularity remained stable for 87% of packages, though synchronized releases saw an 11.36% increase in forks. While contributions increased in some cases (e.g., 13.59% in NPM → PyPI), others saw a decline in commit activity. Overall, cross-ecosystem deployment does not significantly raise maintenance effort but also does not guarantee increased adoption. Our results show insights towards understanding how deploying to multiple ecosystems may have some benefits.

Human activity recognition (HAR) is necessary for detection of unsafe activity in industrial production, but there are still some issues that need to be solved, such as limited data in different scenarios and the lack of a unified model for different situations. Therefore, a novel meta-federated learning framework with distillation of activation boundaries (AB) is proposed, in which a federation is viewed as a meta-distribution and all federations work together without a central server. Specifically, the personalized model from the previous federation serves as the teacher model for the next federation, where general knowledge is extracted by AB knowledge distillation, the personalized knowledge is acquired through local training, and a high-quality model is obtained for the current federation by dynamically fusing general knowledge and personalized knowledge. To evaluate the effectiveness and superiority of the proposed framework, experiments were conducted on one popular HAR datasets (PAMAP2) and a chemical scenario dataset (WACID) constructed by our laboratory. The experimental results show that our proposed framework outperforms the state-of-the-art methods with fewer communication costs, achieving the recognition accuracies of 91.23% and 95.66% on the PAMAP2 dataset and WACID dataset, respectively.

Several works recently focus on monitoring air quality of critical areas using sensors attached to buses. They aim to monitor the maximum number of critical areas using a limited number of sensors. In practice, we may want to have information for all critical areas. We work on the problem of covering all the areas using the minimum number of sensors in this work. We show that, even when the bus routes are not pre-defined, the problem is log-APX-hard and is significantly harder than the problem of the previous works. We develop two algorithms for the case that the routes are pre-defined. Those algorithms include a fixed parameter tractability and a greedy algorithm. Next, we show an NP-completeness reduction for a special case of the problem and propose a 2-approximation algorithm for it. Our experiment results show that, although we usually give the similar number of sensors as the algorithm in the previous works, our algorithms have a shorter computation time than the classical greedy algorithm.

The composition model in ROS 2 enables multiple nodes to run within a single process, reducing the overhead of inter-process communication (IPC). However, this architecture introduces memory safety and concurrency challenges due to a shared address space and a common Executor. Existing tracing tools lack the granularity to detect node-level runtime anomalies in such settings. We present C_OMPS_HIELD, a system that extends ROS2Trace to enable intra-process node-level tracing and misbehavior detection. C_OMPS_HIELD combines static analysis with enhanced runtime tracing to identify temporal anomalies and concurrency-related performance issues, such as Executor monopolization and prolonged callback execution. Evaluation on a ROS 2 composition application shows that C_OMPS_HIELD effectively detects such issues with low overhead.

Unlike conventional speech-based depression detection (SDD), cross-elicitation SDD presents a more challenging task due to the differing speech elicitation conditions between the labeled source (training) and unlabeled target (testing) speech data. In such scenarios, a significant feature distribution gap may exist between the source and target speech samples, potentially reducing the detection performance of most existing SDD methods. To address this issue, we propose a novel deep transfer learning method called the Deep Elicitation-Adapted Neural Network (DEANN) in this letter. DEANN aims to directly learn both depression-discriminative and elicitation-invariant features from speech spectrograms corresponding to different elicitation conditions using two weight-shared Convolutional Neural Networks (CNNs). To achieve this, the CNNs are first endowed with depression-discriminative capability by establishing a relationship between the source speech samples and the provided depression labels. Subsequently, a well-designed constraint mechanism, termed Bidirectional Sparse Reconstruction, is introduced. This mechanism ensures that source and target speech samples can be sparsely reconstructed by each other at the same feature layer of both CNNs, allowing the learned features to maintain adaptability to changes in speech elicitation conditions while preserving their original depression-discriminative capability. To evaluate DEANN, we conduct extensive cross-elicitation SDD experiments on the MODMA dataset. The experimental results demonstrate the effectiveness and superiority of the proposed DEANN in addressing the challenge of cross-elicitation SDD compared to many existing state-of-the-art transfer learning methods.

An interconnection network is an inevitable component in a parallel computer. It offers communication capabilities in the parallel machine, affecting performance issues of parallel computation. Thus, various discussions are being made from a variety of aspects to reduce the communication cost and to improve the performance. This paper addresses the packet scheduling problem, which is a promising method for improving the performance, in the collective communication. Our preceding work has proposed a unique optimization method Lopit (lazy optimization of packet injection timing). This paper extends the method by introducing a group nature in collective communication situations and proposes a new method G-Lopit (grouped Lopit). Evaluation results in our interconnection network simulator reveal the significant effectiveness of the proposed method. The G-Lopit method outperforms the traditional GA and the preceding Lopit methods. It improves the performance of collective communication at most 1.18 times from the Lopit method in 32×32 2D-torus network with bcmp traffic. In comparison with unoptimized situations, it achieves at most 1.73 times improvement in the shfl traffic.

In official promotional videos and media press, the malicious cut-out editing of content to misrepresent the speaker's intended meaning has long been recognized as a problem of misinformation/ disinformation. This issue of selective cut-out editing has gained increasing attention, particularly in the context of the growing need for fact-checking mechanisms. Consequently, there is a strong demand for technical approaches to counteract propaganda strategies that exploit partial multimedia content extraction. However, cut-out editing is also an essential technique for compressing lengthy content into highlights or summaries, making outright regulation of such actions impractical. This study proposes a framework that applies cut-and-paste editing to official video releases, permitting clipping some segments and concatenating them while preventing malicious edits without undermining editorial discretion. By employing semantic analysis to identify potentially exploitable segments and their contextual or meaning-based relationships, the approach restricts intentional misrepresentative cut-and-paste editing. The resulting restricted version is then published as the original content, offering a defensive strategy against such manipulations.

This research addresses the critical information asymmetry between wellness service providers and travelers in the rapidly expanding wellness tourism sector, which affects informed decision-making in one of the fastest-growing markets. Through computational analysis of 430 wellness program descriptions and 6,436 customer reviews from 165 wellness centers in Thailand, we identify significant misalignments between available information and traveler needs. Our sentiment analysis reveals a strong negative correlation between information gaps and customer satisfaction, while supervised classification with 90% accuracy categorizes programs into six wellness dimensions. The resulting integrated wellness program formation introduces essential, recommended, specialized information tiers validated through K-means clustering that revealed eight natural information patterns transcending conventional wellness categorizations. By highlighting critical decision-making factors like health warnings which are currently missing from nearly 90% of descriptions and skill requirements, while respecting cultural differences in wellness concepts, our formation strikes a balance between completeness and clarity. This research establishes evidence-based guidelines for organizing wellness information that significantly improves traveler decision support, enabling better-informed choices across the diverse landscape of wellness tourism.

Deep neural networks (DNNs) remain vulnerable to adversarial attacks that cause misclassification when specific perturbations are added to input images. This vulnerability also threatens the reliability of DNN-based monocular depth estimation (MDE) models, making robustness enhancement a critical need in practical applications. To validate the vulnerability of DNN-based MDE models, this study proposes a projection-based adversarial attack method that projects perturbation light onto a target object. The proposed method employs physics in-the-loop (PITL) optimization—evaluating candidate solutions in actual environments to account for device specifications and disturbances—and utilizes a distributed covariance matrix adaptation evolution strategy. Experiments confirmed that the proposed method successfully created adversarial examples that lead to depth misestimations, resulting in parts of objects disappearing from the target scene.

Steganography, the art of information hiding, has continually evolved across visual, auditory and linguistic domains, adapting to the ceaseless interplay between steganographic concealment and steganalytic revelation. This study seeks to extend the horizons of what constitutes a viable steganographic medium by introducing a steganographic paradigm in robotic motion control. Based on the observation of the robot's inherent sensitivity to changes in its environment, we propose a methodology to encode messages as environmental stimuli influencing the motions of the robotic agent and to decode messages from the resulting motion trajectory. The constraints of maximal robot integrity and minimal motion deviation are established as fundamental principles underlying secrecy. As a proof of concept, we conduct experiments in simulated environments across various manipulation tasks, incorporating robotic embodiments equipped with generalist multimodal policies.

Container-based virtualization enables fast and lightweight execution of independent virtual environments. Docker, the de facto standard for container-based virtualization software, has been widely used in recent years. In Docker, virtual environments are created by building a text file called a Dockerfile. However, docker builds can cause build errors, especially due to complex dependencies between Dockerfile instructions.

To address this issue, this study proposes an automatic build error repair method focusing on dependencies. The proposed method first identifies dependencies that induce build errors and then automatically repairs them. For identifying dependencies that induce build errors, sequential pattern mining is applied to Dockerfiles in the dataset. The build error rate of Dockerfiles containing the obtained sequential patterns is then calculated, and Dockerfiles with a high build error rate are treated as repair candidates. Next, repair rules are created for the identified candidates, and an automatic build error repair process is applied.

As a result, 4,693 dependencies that induce build errors were extracted, and 28 repair rules were generated based on the extracted dependencies. Among the 4,554 Dockerfiles in the dataset, the proposed method successfully repaired 328 files, which is 25.7% more than the baseline method so called Shipwright. The proposed method also repaired 195 build errors that Shipwright could not address. Furthermore, a comparison between the repairs performed by the proposed method and actual project changes showed that 65 out of 192 (33.5%) collected project modifications matched the repairs made by the proposed method.

This paper explores a practical means to employ Data Plane Development Kit (DPDK), a kernel-bypassing framework for packet processing, in resource-limited multi-tenant edge data centers. Traditional virtual CPU (vCPU) schedulers are not compatible with the event detection model of DPDK, because each DPDK-applied VM (virtual machine) monopolizes one physical CPU (pCPU) for NIC register polling. Toward edge data center providers, this paper presents a new vCPU scheduling policy named Polling vCPU Consolidation (PvCC). PvCC enables DPDK-applied VMs to be consolidated on dedicated pCPUs by adopting microsecond-scale time slices. Along with this, PvCC provides vCPU scaling API that enables to scale up/down vCPUs of VMs at runtime. Our experiments show that PvCC allows consolidated DPDK-applied VMs to achieve low serving latencies, and our vCPU scaling API allows us to adjust CPU resource assignment according to the incoming request rate. It also enables to assign spare pCPUs to VMs executing non-latency-sensitive tasks.

The Video-based Dynamic Mesh Coding (V-DMC) standard is a cutting-edge technology for compressing dynamic mesh data. V-DMC realizes parallel and partial decoding by introducing submesh frameworks, where dynamic meshes are segmented and processed independently. However, V-DMC may introduce submesh boundary errors due to misalignments in the existence or coordinates of vertices. Here, submesh boundary errors are defined as the misalignments of vertex coordinates and connectivity between adjacent submesh boundaries. These errors can create holes, thereby degrading both the objective and subjective quality of the decoded dynamic meshes. To minimize boundary errors and enhance coding performance, we propose a two-stage boundary error correction method integrated into V-DMC's preprocessing and encoding/decoding processes. Specifically, the first stage involves rearranging the preprocessing order to minimize boundary errors, and the second stage fills holes using boundary information. Experimental results demonstrate that the proposed method effectively minimizes boundary errors in V-DMC decoded meshes, significantly improving both objective and subjective quality compared to the V-DMC reference software. The D1 BD-Rate gain, indicative of coding efficiency, averages -16.3% for all intra and -25.9% for random access conditions.

With the spread of smartphones and AI, enterprises need to accelerate innovation to respond quickly to changes in customer needs. In innovation, new core technologies are embodied in software, evolved, and ultimately enhanced to customer value. However, as innovation activities spread across multiple organizations, there was a risk that the final quality requirements would be over- or underdeveloped, causing problems involving customers, or that development costs would not be optimized. In this paper, we propose a self-assessment method for R&D organizations at the center of innovation to accurately grasp the initial quality requirements of core technologies and hand them as software products over to other organizations to surely realize quality requirements that satisfy customer values. The proposed method was applied to all R&D organizations from 2014 as a part of software risk management measures, and it was confirmed that it contributed to avoidance of a critical incident resulting in a service outage and optimization of development cost.

Diffusion models are generative models that generate images, videos, and audio through learning samples and have attracted attention in recent years. In this paper, we investigate whether a diffusion model is resistant to membership inference attacks, which evaluate the privacy leakage of a machine learning model. We primarily discuss the diffusion model from the standpoints of comparison with a generative adversarial network (GAN) as a conventional model and hyperparameters unique to the diffusion model, such as timesteps, sampling steps, and sampling variances. We conduct extensive experiments with the denoising diffusion implicit model (DDIM) as a diffusion model and the deep convolutional GAN (DCGAN) as a GAN on the CelebA and CIFAR-10 datasets in both white-box and black-box settings and then show that the diffusion model is comparable to GAN in terms of resistance to membership inference attacks. Next, we demonstrate that the impact of timesteps is significant and that the intermediate steps in a noise schedule the most vulnerable to the attack. We also found two key insights through further analysis. First, we identify that DDIM is more vulnerable to the attack when trained with fewer samples even though it achieves lower Frechet inception distance scores than DCGAN. Second, sampling steps in hyperparameters are important for resistance to the attack, whereas the impact of sampling variances is negligible.

With the rapid development of autonomous driving vehicles, it has become clear that the automotive domain also needs to embrace the transformation for new mobility services. To improve the reliability on autonomous driving systems, recent works focus on generating natural language explanations (captions) for recorded driving scenes, which are expected to provide grounds for driving risk analysis. To evaluate captioning performance for challenging risky driving scenes, the DRAMA benchmark dataset was recently introduced, where each video is annotated with question-answer (QA) pairs identifying the risk object and a caption summarizing the risk event. However, existing methods typically generate captions without explicitly modeling intermediate reasoning processes, limiting interpretability and making it challenging to pinpoint specific sources of captioning errors, such as incorrect localization or inaccurate inference of risk causes. This lack of transparency also hinders the analysis of model hallucinations and restricts further improvements in model reasoning. To address this limitation, we propose a two-stage chain-of-thought video captioning framework using pretrained models, which first use QA pairs to finetune a model generate rationales regarding to the target object, which are then served as intermediate reasoning steps for finetuning the captioning model. Additionally, we introduce a visual prompting approach by highlighting risk objects with red bounding boxes to explicitly guide model attention. The experimental results show that our rationale-based approach outperforms the existing state-of-the-art methods by 13.5% in BLEU-4 and 11% in CIDEr scores on the DRAMA dataset. Furthermore, our analysis of the generated rationales reveals that the pretrained model still struggles to reason about the correct cause of the risk, suggesting potential directions for future work.

We have developed a Scalable CI/CD Pipeline to address internal challenges related to Japan 2025 cliff problem, a critical issue where the mass end of service life of legacy core IT systems threatens to significantly increase the maintenance cost and black box nature of these system also leads to difficult update moreover replace, which leads to lack of progress in Digital Transformation (DX). If not addressed, Japan could potentially lose up to 12 trillion yen per year after 2025, which is 3 times more than the cost in previous years. Asahi also faced the same internal challenges regarding legacy system, where manual maintenance workflows and limited QA environment have left critical systems outdated and difficult to update. Middleware and OS version have remained unchanged for years, leading to now its nearing end of service life which require huge maintenance cost and effort to continue its operation. To address this problem, we have developed and implemented a Scalable CI/CD Pipeline where isolated development environments can be created and deleted dynamically and is scalable as needed. This Scalable CI/CD Pipeline incorporate GitHub for source code control and branching, Jenkins for pipeline automation, Amazon Web Services for scalable environment, and Docker for environment containerization. This paper presents the design and architecture of the Scalable CI/CD Pipeline, with the implementation along with some use cases. Through Scalable CI/CD, developers can freely and safely test maintenance procedures and do experiments with new technology in their own environment, reducing maintenance cost and drive Digital Transformation (DX).

To realize low-latency spatial transmission system for immersive telepresence, there are two major problems: capturing dynamic 3D scene densely and processing them in real time. LiDAR sensors capture 3D in real time, but produce sparce point clouds. Therefore, this paper presents a high-speed LiDAR point cloud densification method to generate a dense 3D scene with minimal latency, addressing the need for on-the-fly depth completion while maintaining real-time performance. Our approach combines multiple LiDAR inputs with high-resolution color images and applies a joint bilateral filtering strategy implemented with a CNN-inspired architecture. Experiments demonstrate that the proposed method produces dense depth maps at full HD resolution in real time (30 fps), which is over 15x faster than a recent training-based depth completion approach [1]. The resulting dense point clouds exhibit accurate geometry without multi-view inconsistencies or ghosting artifacts.

This letter describes a training-free 3D semantic segmentation method using virtual cameras and a 2D foundation model guided by language prompts. Aggregating multi-view predictions via weighted voting achieves accuracy comparable to supervised methods and supports open-vocabulary recognition without requiring annotated 3D data or paired RGB images.

Third-party libraries are increasingly used in software development. With new versions released, many conflicts due to API breaking may be introduced. Therefore, semantic versioning with a set of rules and requirements is proposed to inform users about incompatibilities and other changes in a new release. Simply speaking, major, minor, and patch version numbers should be updated when breaking, non-breaking, and internal changes are made, respectively. However, many third-party libraries do not follow semantic versioning principles, bringing many efforts to adapt new versions of libraries. Hence developers are often unwilling to update dependencies. In this paper, we first retrospectively investigate the three types of existing techniques that can be used in automatic versioning, a task for more reliable versions: 1) rule-based, considering keywords only, 2) machine learning-based, concerning many aspects of features, and 3) source code analysis for semantic versioning compliance. Having found the limitation of these approaches, we propose a new and simple approach AutoVer, which can capture the intent of developers through commit messages. Specifically, we treat the tokens and phrases in the commit messages as features and train AutoVer using XGBoost, a well-known machine learning method that performs well in many classification tasks. The evaluation results show that AutoVer outperforms our investigated approaches in terms of many metrics. Specifically, the major, minor and major type F1-Scores of AutoVer are 0.889, 0.992 and 0.998, respectively. We also comprehensively investigate many settings that affect the performance of AutoVer, including choice of machine learning model, number of features and keywords in commit messages as well as provide practitioners and researchers with some implications for future studies, e.g., writing clear commit messages for better understanding the intent of making changes, and combining semver-compliance checking and automatic versioning.

Cross-domain recommendation (CDR) leverages knowledge from a source domain to improve recommendation accuracy in a target domain with scarce data. Recent approaches attempt to disentangle user preferences into shared and domain-specific representations. However, most existing methods rely solely on simple user-item interaction information, making effective disentanglement challenging. Methods that employ contrastive learning to separate representations optimize relative distances, which may fail to suppress semantic overlap between representations and lead to inappropriate knowledge transfer. We propose a novel CDR method that integrates multimodal features and introduces a cosine similarity-based regularization term to suppress correlations between shared and domain-specific representations, ensuring each representation captures semantically distinct information. Our approach features: (1) rich representation learning using diverse modalities, (2) direct orthogonality constraints between representations via cosine similarity, and (3) explicit directional correlation suppression in the representation space. Experiments on large-scale Amazon datasets demonstrate our method outperforms state-of-the-art CDR approaches by an average of 3.7% over the best baseline in each domain. The code is available at https://github.com/meruemon/MMCDR.

A 360 degrees panoramic image can be acquired by multiple cameras which cover the surrounding scenes as a camera cluster. The mostly popular and convenient approach is to use a pair of fisheye cameras which point opposite directions with a wider field of view more than hemisphere, respectively. To synthesize a 360 degrees panoramic image, traditional methods need to calibrate the intrinsic parameters of each fisheye camera and estimate the relative pose (extrinsic parameters) of the pair of fisheye cameras. To acquire upright panoramic image, the inclination correction is further needed. Until now, camera parameters and inclination correction are coped with differently. In this paper, we propose a novel deep learning-based method of generating a upright panoramic image directly from a pair of fisheye cameras. That is, we develop a neural network to generate a 360 degrees panoramic image from a pair of inclined fisheye images. First, we collected a comprehensive dataset specifically designed for this task. Then, we constructed a neural network to achieve this task, directly generating upright panoramic images from a pair of inclined fisheye images. The experimental results shown the effectiveness of the proposed method, also.

The Light Up puzzle is one of the Nikoli's puzzles, and its NP-completeness has already been shown [8]. In this paper, we introduce a new variant of Light Up that naturally arises from its connection with the n-Queens problem, and discuss the problem of deciding whether it has a solution for this Light Up variant under various restrictions. This variant extends the directions of light from four (up, down, left, right) to eight, including diagonal directions, and this extension corresponds to the chess queen's movement in the n-Queens problem [2]. In the original Light Up puzzle, the numbers 0, 1, 2, 3, and 4 represent the number of light bulb placements. However, when considering the extension of the light irradiation direction, the number of light bulb placements is limited to 0, 1, and 2. In this paper, we analyze the computational complexity of the Light Up variant that limits the number of light bulb placements to only one type, and prove its NP-completeness using a reduction from Circuit-SAT, specifically for the cases where the number of light bulb placements is restricted to 0 or 1. These results also show that all problems are NP-complete except for the case where the number of light bulb placements is exactly 2.

The dispersive art gallery problem is to find a guard set for a polygon such that every pair of guards maintains the maximum possible distance from each other. In this paper, we study a chromatic variant of this problem, where each guard is assigned one of k distinct colors. The chromatic dispersive art gallery problem is to find a guard set for a polygon such that every pair of guards having the same color are placed as far apart from each other as possible. We study the decision version of this problem when the instance is a polyomino, which is the union of connected unit squares. In this paper, it is shown that determining whether there exists an r-visibility guard set for a polyomino with holes such that every two guards with the same color are placed at a distance of at least 6 is NP-complete when the number of colors is k = 2. Here, two points are r-visible if the smallest axis-aligned rectangle containing them lies entirely within the polyomino.

Semi-supervised video object segmentation (SVOS) is a challenging task that uses an initial frame mask to predict the segmentation of target objects in subsequent frames. Recently, various VOS methods have combined matching-based transductive inference with online inductive learning to capture more precise spatiotemporal information, thereby enhancing segmentation accuracy. However, while these methods improve feature extraction capabilities, they still fail to adequately address the full fusion of different features for more efficient feature utilization. To address the issue of low efficiency in feature fusion utilization in SVOS, we propose an adaptive multi-feature fusion method in this letter. This method proposes a Foreground-Background Multi-feature Encoder to effectively enhance feature diversity and uses a Multi-feature Fusion Module to dynamically integrate spatiotemporal cues from both the foreground and background. For different segmentation targets, the method employs a Feature Fusion Reader to autonomously select and adaptively fuse multiple foreground-background features, thereby achieving inter-feature optimization and significantly improving target-specific fusion efficiency. Extensive experiments on DAVIS 2017 and large-scale YouTube-VOS 2018/2019 datasets demonstrate that our proposed method achieves state-of-the-art performance.

Annotated corpora play a crucial role in supervised machine learning, particularly in natural language processing (NLP); however, labels assigned by humans are not always reliable. Recent studies have predominantly employed the k-fold cross-validation method, which involves training machine learning models to predict data labels and detect annotation errors. Although increasing the number of folds or performing multiple validation cycles with different training subsets can improve error detection accuracy, these methods often result in considerable computational cost. In this study, we propose the Annotation Error Candidate Cross-Validation method, which selectively uses data with low error probabilities solely for training, while repeatedly evaluating high-error-probability data with various trained models. Our experimental results demonstrate that the proposed method accurately identifies deliberately inserted label errors in NLP tasks. Furthermore, to support the re-annotation process, we investigate the extent to which error probabilities should be reviewed to optimize annotation efficiency.

This study demonstrates the effectiveness of Domain-Adaptive Pre-Training (DAPT) on pre-trained language models for cyber-security text mining. Experiments using various versions of BERT and RoBERTa confirm significant improvements in domain-specific tasks, particularly in identifying security-related concepts and terminology, and reveal that DAPT's effectiveness varies depending on the label type.

This study presents design-rule compliant optical system guidelines for laser speckle authentication, clarifying key parameter constraints. We reveal the relationship between sampling and speckle size, and experimentally show enhanced robustness against displacement. Quantitive analysis using false acceptance and false rejectance rates confirms reliable performance of our design-rule compliant system.

We consider a simple graph G = (V, E) with vertex set V and edge set E. Let G-u be a subgraph induced by the vertex set V ＼ {u}. The distance δ_G (x, y) is the shortest path's length between the vertices x and y in G. The vertex u ∈ V is a hinge vertex if there exist two vertices x, y ∈ V ＼ {u} such that δ_G-u (x, y) > δ_G (x, y). Let U be a set consisting of all hinge vertices of G. The open neighborhood of u is denoted by N (u). We defined the detour degree of u as det (u) = max{ δ_G-u (x, y) | δ_G-u (x, y) > δ_G (x, y), x, y ∈ N (u)} for u ∈ U. The detour hinge vertex problem aims to determine the hinge vertex u that maximizes det (u) in G. In this study, we propose an efficient algorithm for solving the detour hinge vertex problem on trapezoid graphs that runs in O (n²) time, where n is the number of vertices in the graph.

Secure multi-party computation (MPC) allows participating parties to jointly compute a function over their inputs while keeping them private. In particular, MPC based on additive secret sharing has been widely studied as a tool to obtain efficient protocols secure against a dishonest majority, including the important two-party case. In this paper, we propose a two-party protocol for an exponentiation functionality based on an additive secret sharing scheme. Our proposed protocol aims to securely compute a public base exponentiation a^x mod p for an odd prime p, where the exponent x ∈ ℤ_p is a (shared) secret and the base a ∈ ℤ_p is public. Our protocol is based on a new simple but efficient approach involving quotient transfer that allows the parties to perform the most expensive part of the computation locally, but requires the base a ∈ ℤ_p to be a quadratic residue. To address scenarios where the base does not fulfill this, we combine our exponentiation protocol with a new efficient modulus conversion protocol which might be of independent interest. Even taking into account a potential modulus conversion, our exponentiation protocol only requires 3 rounds and 4 invocations of multiplication.

Title-overlaid images are useful as thumbnails on social media, where users prefer concise information to share and watch contents. Focusing on food contents, we aim to support creation of attractive title-overlaid food images to attract viewers' attentions. This paper first analyzes the effect of font styles of the title on the attractiveness of title-overlaid images via preference experiments, and creates a dataset. Next, we design a prototype model of attractive font selection for a food image and its title. Its effectiveness is demonstrated through experiments on the created dataset.

We introduce innovative design concepts to address hard segmentation cases in complex image backgrounds and small training data sizes. We define (1) the shape feature description, including the medial surface, mask segmentation, and contour shape, which characterize articular cartilage shape, and establish a cartilage segmentation network to achieve multi-task consistency; and (2) the shape prior description, representing the shape distribution of articular cartilages, and establish a neural network based on this description. We incorporate the shape prior network into the multi-task consistency segmentation network. This results in a deep learning framework with high accuracy and strong generalization, guided by shape feature description and constrained by shape prior description. Our framework handles difficult cases with low contrast, ambiguous boundaries, deformed portions, and touching cartilages. The effectiveness of our method is demonstrated on two public knee image datasets and one clinical hip image dataset, where our approach shows increased segmentation accuracy compared to other state-of-the-art methods. Furthermore, its generalization is demonstrated for a subset of the BTCV dataset focusing on three specific structures: the aorta, the inferior vena cava, and the portal and splenic veins.

In the past few years, research on lightweight block ciphers as security ciphers in the Internet of Things (IoT) has attracted considerable attention in cryptography. In this paper, we present an improved framework for neural distinguishers in lightweight SPN block ciphers suitable for IoT, focusing on two aspects: training data format and neural network structure. First, we analyze the nature of the SPN round function, then divide it into three cases to apply data augmentation. Second, we generate training data samples using three dimensions and construct neural networks using two-dimensional convolution. Finally, we validate the advantages of the improved framework on the SKINNY family and MIDORI family with higher accuracy and achieve a breakthrough in the number of rounds.

The Burrows-Wheeler Transform (BWT) is a core technology in many modern compression and bioinformatics applications. Constructing the BWT for dynamically growing string collections remains a challenge. The existing optimal BWT (optBWT) construction algorithm for string collections can significantly reduce the number of BWT runs r for a given string collection. However, it requires reconstructing the entire BWT when new strings are added. To address this issue, this paper proposes an online BWT construction algorithm based on dynamic insertion interval — onptBWT (Online Computation of Near-Optimal BWT). This algorithm requires onlyO (r log n) bits of space andO (m log r) time (where n is the dataset length, mis the length of the newly added string, and r is the number of runs) for each new string added to produce a BWT that is near-optimal by only a small margin compared to optBWT in terms of runs r. Experimental results show that, across seven real-world genomic datasets, the average number of runs produced by onptBWT is only 1.41% higher than that of optBWT, outperforming six other BWT construction algorithms. In scenarios where the newly added strings are significantly smaller than the original string collection, onptBWT achieves faster construction speed and lower peak memory usage compared to the offline optBWT algorithm. Source code is available at https://github.com/xiaoYu0103/onptBWT.git.

Hebbian learning limits Hopfield network storage capacity (pattern-to-neuron ratio around 0.14). We propose Kernel Logistic Regression (KLR) learning. Unlike linear methods, KLR uses kernels to implicitly map patterns to high-dimensional feature space, enhancing separability. By learning dual variables, KLR dramatically improves storage capacity, achieving perfect recall even when pattern numbers exceed neuron numbers (up to ratio 1.5 shown), and enhances noise robustness. KLR demonstrably outperforms Hebbian and linear logistic regression approaches.

Oredango puzzle, one of the pencil puzzles, was originally created by Kanaiboshi and published in the popular puzzle magazine Nikoli. In this paper, we show NP- and ASP-completeness of Oredango by constructing a reduction from the 1-in-3SAT problem. Next, we formulate Oredango as a 0-1 integer-programming problem, and present numerical results obtained by solving Oredango puzzles from Nikoli and PuzzleSquare JP using a 0-1 optimization solver.

We examine the security of AES-based authenticated encryption schemes including the AEGIS family, Tiaoxin-346, Rocca and Rocca-S. In existing work, Takeuchi et al. evaluated the security against differential attacks for the initialization phase of AEGIS and Rocca and estimated the lower bounds for the number of active S-boxes by byte-wise search. Similarly, Shiraya et al. evaluated the security against differential attacks for the initialization phase of Tiaoxin-346. However, the byte-wise evaluation might underestimate the lower bounds for the number of active S-boxes as it might include invalid characteristics. In this paper, we conduct a bit-wise evaluation of the AEGIS family, Tiaoxin-346, Rocca, and Rocca-S to obtain exact lower bounds for the number of active S-boxes by the boolean satisfiability problem (SAT) tools. Besides, we derive the optimal differential characteristics in the initialization phase. As a result, we update the lower bounds for the number of active S-boxes in some rounds for each targets and derived the probability of optimal differential characteristics for the first time, which allowed us to find longest differential distinguisher of initialization phase on AEGIS family, Tiaoxin-346, Rocca and Rocca-S. As far as we known, our results are the first bit-wise distinguishing attacks on initialization phases of AEGIS, Tiaoxin-346, Rocca, and Rocca-S.

Zoom Mail, an email service provided by Zoom Video Communications, features a proprietary end-to-end encryption (E2EE) scheme, which is detailed in a whitepaper available on GitHub. To date, there has been no detailed discussion or third-party evaluation of the security of the E2EE implementation in Zoom Mail. In this paper, we conduct a comprehensive security analysis of Zoom Mail's E2EE. Specifically, we establishes four types of adversary models: insiders, outsiders, TO/CC (carbon copy) members, and BCC (blind carbon copy) member. We focus on three security goals: confidentiality, integrity, and authenticity. Based on these adversary models and security goals, we present the results of security evaluation of Zoom Mail's E2EE for the first time.

A blockchain is a distributed ledger that allows users to exchange information without a centralized authority. This technology enables users to send and receive tokens among other applications, such as transactions, product management, and elections. It is possible to send data and tokens inside a single blockchain, but a method to efficiently share the data and tokens among different blockchains has not yet been con structed. Cross-chain communication, the focal point of several recent research efforts, is a scheme for sending data or tokens among different blockchains. In existing studies, a trusted third party (TTP) is used to ensure fair rates of token exchange among different blockchains. However, because blockchains are originally designed with a policy that does not incorporate the use of TTPs, the fair exchange rate should not be determined by TTPs, but rather by the market price of tokens among users. When exchange rates are determined from quotes among users, the preferred scheme is to determine the exchange rate offered by many users as an auction. Here, some existing cross-chain communication systems use smart contracts that automatically execute arbitrary processes on the blockchain. However, such schemes require a gas fee each time a smart contract is executed. Thus, implementing an auction scheme that determines the fair exchange rate among different blockchains would necessitate each user to pay a fee for each new token offered, which would result in high gas fees. In this study, we propose a scheme to deter mine exchange rates from quotes among users with a relatively low gas fee. Using a first-price sealed-bid auction and commit ment scheme, the user with the highest token value can be identified without revealing the other users' token offer values. In our scheme, the largest token value among users is determined as the exchange rate using an external Smart Contract (SC) instead of a TTP. We further modify the existing insert key-value com mitment scheme to aggregate the commitment values of token offers. Our scheme is based on the generalized RSA assumption. By proving that it satisfies the key-binding property, we prove that the token sender cannot act maliciously. We further implement the proposed scheme and demonstrate that the gas fees and data space required to implement the proposed scheme are practically feasible.

This letter presents a vertical-based adaptation of SPADE on Spark that significantly minimizes inter-worker communication. We achieve up to 6.2 × speedup over Spark MLlib's PrefixSpan, enabling more efficient sequential pattern mining with minimal data movement and strong performance in distributed environments.

In this paper, we propose a method that combines an anisotropically weighted version of the directional relative total variation measure (AW-dRTV) with a pixel-wise Intensity and Scale Adjustable Edge-Preserving Smoothing filter (ISES filter). This approach enables effective removal of high-contrast textures with low computational cost while preserving important structural edges.

Precipitation is a crucial component of the natural water cycle, and inadequate timeliness and precision in precipitation prediction can result in agricultural losses, traffic disruptions, flood catastrophes, and even threats to human life. Consequently, precipitation prediction is a key problem in the domain of meteorology. However, the current methodologies pay close attention to the explicit spatial connections of precipitation regions while neglecting the implicit spatial connections over time. There are often challenging for traditional convolutional neural networks and graph neural networks to capture, leading to inaccurate spatial regions and poor timeliness of model predictions. To resolve this problem, we propose a Dynamic spatial-temporal graph prediction model for short-term precipitation (Dst-pred), which dynamically explores implicit connections among meteorological stations in the target region through graph neural networks and constructs dynamic spatial-temporal graphs to predict precipitation in the region. We have verified our Dst-pred model on our proprietary precipitation dataset from Guangxi Province, China, and the ERA5-Land dataset, and it can extract the implicit spatial connections between individual stations from the precipitation data of meteorological stations. The precipitation process capture of our model enhances the timeliness and accuracy of nowcasting precipitation prediction with the best performance.