The Financial Action Task Force (FATF) Travel Rule compels each virtual asset service provider (VASP) to exchange originator and beneficiary identities, creating a policy-driven tension between regulatory compliance and user privacy. This yields a concrete requirement: comply with the regulation while ensuring strong privacy and security across domains. However, existing approaches often address only a subset of these requirements in multi-VASP settings (e.g., configurable t-of-n tracing without explicit protection against log-driven cross-domain linking or Δ-revocation safety). We present PAAS, a policy-adaptive traceable anonymity scheme for cross-domain networks. PAAS enforces bridging-log privacy and policy-driven split tracing for incident-tier governance by mapping trace tiers to a hierarchical, role-weightedt-of-nthreshold structure, thereby separating pseudonym linkage from real-world identity disclosure (link ≠ identity), and bounds revocation-bypass risk within a latency window Δ under a Δ-revocation model with an ephemeral challenge–response bridging protocol. Together, these mechanisms deliver unlinkable anonymity, accountable tracing, and safe anonymous roaming across domains. We analyze the security of PAAS under standard cryptographic assumptions, establishing robust policy-adaptive anonymity and policy-governed accountability; our performance evaluation reports service-phase computation latency of 0.22 ms (existing pseudonym) or 6.12 ms (with bridging issuance) and throughput of ≈49.5 TPS (i.e., requests/s) under 100 concurrent users, outperforming the evaluated baselines in terms of service-phase latency and throughput, while maintaining low-latency computation for bridging and roaming (0.7–6.6 ms).

The significance of machine-learning approaches in the healthcare domain has grown rapidly owing to the existence of enormous amounts of data and well-established simulation models and algorithms. The digitization of health-related data, as well as rapid technological advancements are accelerating the development and application of machine learning in healthcare, particularly in precision medicine. The ultimate goal of precision medicine is to provide personalized medicine, which requires tailoring medical decisions to each patient based on their projected disease response. In this study, we propose a cluster-applied deep reinforcement learning-based type 2 diabetes treatment recommendation model based on the electronic health records of South Koreans. The purpose of applying a clustering algorithm is to group patients who are in a similar state, to boost the performance of deep reinforcement learning, build a more realistic treatment recommendation model to support clinicians, and develop expert systems in the field of healthcare. The proposed model demonstrated significant performance by decreasing diabetes-related medical checkup measurements. Furthermore, the proposed model delivered high-quality performance when compared with existing reinforcement-learning methods. Finally, the recommendation outcomes of the proposed model were validated against real-life prescriptions to ensure the accuracy of the findings.

From a streaming video, online action detection aims to identify actions in the present. For this task, previous methods use recurrent networks to model the temporal sequence of current action frames. However, these methods overlook the fact that an input image sequence includes background and irrelevant actions as well as the action of interest. For online action detection, in this paper, we propose a novel recurrent unit to explicitly discriminate the information relevant to an ongoing action from others. Our unit, named Information Discrimination Unit (IDU), decides whether to accumulate input information based on its relevance to the current action. This enables our recurrent network with IDU to learn a more discriminative representation for identifying ongoing actions. In experiments on two benchmark datasets, TVSeries and THUMOS-14, the proposed method outperforms state-of-the-art methods by a significant margin. Moreover, we demonstrate the effectiveness of our recurrent unit by conducting comprehensive ablation studies.

We propose a simple yet efficient anchor-free instance segmentation, called CenterMask, that adds a novel spatial attention-guided mask (SAG-Mask) branch to anchor-free one stage object detector (FCOS) in the same vein with Mask R-CNN. Plugged into the FCOS object detector, the SAG-Mask branch predicts a segmentation mask on each box with the spatial attention map that helps to focus on informative pixels and suppress noise. We also present an improved backbone networks, VoVNetV2, with two effective strategies: (1) residual connection for alleviating the optimization problem of larger VoVNet \cite{lee2019energy} and (2) effective Squeeze-Excitation (eSE) dealing with the channel information loss problem of original SE. With SAG-Mask and VoVNetV2, we deign CenterMask and CenterMask-Lite that are targeted to large and small models, respectively. Using the same ResNet-101-FPN backbone, CenterMask achieves 38.3%, surpassing all previous state-of-the-art methods while at a much faster speed. CenterMask-Lite also outperforms the state-of-the-art by large margins at over 35fps on Titan Xp. We hope that CenterMask and VoVNetV2 can serve as a solid baseline of real-time instance segmentation and backbone network for various vision tasks, respectively. The Code is available at https://github.com/youngwanLEE/CenterMask.

Moving object detection in a video plays an important role in many vision applications. Recently, moving object detection using appearance modeling based on a convolutional neural network has been actively developed. However, the CNN-based methods usually require the user's supervision of the first frame so that it becomes highly dependent on the training dataset. In contrast, the method of finding a foreground, which models a background occupying a large proportion in an image, can detect a moving object efficiently in an unsupervised manner. However, existing methods based on background modeling in a pan-tilt-zoom (PTZ) camera suffer many false positives or loss of moving objects due to the estimation error of camera motion. To overcome the aforementioned limitations, we propose a moving object detection method for a PTZ camera through two background models. In an unsupervised way, our method builds the two background models that have different roles: 1) a coarse background model for detecting large changes, and 2) a fine background model for detecting small changes. In more detail, the coarse background model builds a block-based Gaussian model, and the fine model builds a sample consensus model. Both models are adaptively updated according to the estimated camera motion in the video recorded by a PTZ camera. Then, each foreground result from two background models is incorporated to fill the moving object region. Through experiments, the proposed method achieves better performance than the state-of-the-art methods and operates in real-time without parallel processing. In addition, we showed the effectiveness of the proposed model through improved results of moving object detection through combination with the latest supervised method.

Recent temporal action proposal generation approaches have suggested integrating segment- and snippet score-based methodologies to produce proposals with high recall and accurate boundaries. In this paper, different from such a hybrid strategy, we focus on the potential of the snippet score-based approach. Specifically, we propose a new snippet score-based method, named Snippet Relatedness-based Generator (SRG), with a novel concept of “snippet relatedness”. Snippet relatedness represents which snippets are related to a specific action instance. To effectively learn this snippet relatedness, we present “pyramid non-local operations” for locally and globally capturing long-range dependencies among snippets. By employing these components, SRG first produces a 2D relatedness score map that enables the generation of various temporal intervals reliably covering most action instances with high overlap. Then, SRG evaluates the action confidence scores of these temporal intervals and refines their boundaries to obtain temporal action proposals. On THUMOS-14 and ActivityNet-1.3 datasets, SRG outperforms state-of-the-art methods for temporal action proposal generation. Furthermore, compared to competing proposal generators, SRG leads to significant improvements in temporal action detection.

We present a novel image editing system that generates images as the user provides free-form masks, sketches and color as inputs. Our system consists of an end-to-end trainable convolutional network. In contrast to the existing methods, our system utilizes entirely free-form user input in terms of color and shape. This allows the system to respond to the user's sketch and color inputs, using them as guidelines to generate an image. In this work, we trained the network with an additional style loss, which made it possible to generate realistic results despite large portions of the image being removed. Our proposed network architecture SC-FEGAN is well suited for generating high-quality synthetic images using intuitive user inputs.

In visual surveillance systems, it is necessary to recognize the behavior of people handling objects such as a phone, a cup, or a plastic bag. In this paper, to address this problem, we propose a new framework for recognizing object-related human actions by graph convolutional networks using human and object poses. In this framework, we construct skeletal graphs of reliable human poses by selectively sampling the informative frames in a video, which include human joints with high confidence scores obtained in pose estimation. The skeletal graphs generated from the sampled frames represent human poses related to the object position in both the spatial and temporal domains, and these graphs are used as inputs to the graph convolutional networks. Through experiments over an open benchmark and our own data sets, we verify the validity of our framework in that our method outperforms the state-of-the-art method for skeleton-based action recognition.

Visual relationship detection is an intermediate image understanding task that detects two objects and classifies a predicate that explains the relationship between two objects in an image. The three components are linguistically and visually correlated (e.g. "wear" is related to "person" and "shirt", while "laptop" is related to "table" and "on") thus, the solution space is huge because there are many possible cases between them. Language and visual modules are exploited and a sophisticated spatial vector is proposed. The models in this work outperformed the state of arts without costly linguistic knowledge distillation from a large text corpus and building complex loss functions. All experiments were only evaluated on Visual Relationship Detection and Visual Genome dataset.

Guaranteeing end-to-end QoS for multimedia traffic in the IP transport network is a challenging issue. Monitoring the one way delay of the multimedia traffic is necessary but no clear solution for the performance monitoring in the core network exists. In the previous work, we introduced an one-way delay monitoring mechanism called DiffProbe. We introduced the basic concept and the implementation details in the MPLS core network. In this paper, we provided an analytical model of the scheme and investigated the performance of the proposed delay measurement mechanism. We evaluate the accuracy of the delay measurement mechanism and the efficiency of the control mechanism with the simulation. The numerical result shows that our proposed algorithm captures the delay performance of the real traffic when the traffic load is high.