Search

4D medical images, which represent 3D images with temporal information, are crucial in clinical practice for capturing dynamic changes and monitoring long-term disease progression. However, acquiring 4D medical images poses challenges due to factors such as radiation exposure and imaging duration, necessitating a balance between achieving high temporal resolution and minimizing adverse effects. Given these circumstances, not only is data acquisition challenging, but increasing the frame rate for each dataset also proves difficult. To address this challenge, this paper proposes a simple yet effective Unsupervised Volumetric Interpolation framework, UVI-Net. This framework facilitates temporal interpolation without the need for any intermediate frames, distinguishing it from the majority of other existing unsupervised methods. Experiments on benchmark datasets demonstrate significant improvements across diverse evaluation metrics compared to unsupervised and supervised baselines. Remarkably, our approach achieves this superior performance even when trained with a dataset as small as one, highlighting its exceptional robustness and efficiency in scenarios with sparse supervision. This positions UVI-Net as a compelling alternative for 4D medical imaging, particularly in settings where data availability is limited. The source code is available at https://github.com/jungeun122333/UVI-Net., Comment: CVPR 2024

Understanding a social network is a fundamental problem in social network analysis because of its numerous applications. Recently, user engagement in networks has received extensive attention from many research groups. However, most user engagement models focus on global user engagement to maximize (or minimize) the number of engaged users. In this study, we formulate the so-called Local User Engagement Maximization (LUEM) problem. We prove that the LUEM problem is NP-hard. To obtain high-quality results, we propose an approximation algorithm that incorporates a traditional hill-climbing method. To improve efficiency, we propose an efficient pruning strategy while maintaining effectiveness. In addition, by observing the relationship between the degree and user engagement, we propose an efficient heuristic algorithm that preserves effectiveness. Finally, we conducted extensive experiments on ten real-world networks to demonstrate the superiority of the proposed algorithms. We observed that the proposed algorithm achieved up to 605% more engaged users compared to the best baseline algorithms.

With the proliferation of mobile technology and IT development, people can use social network services at any place and anytime. Among many social network mining problems, identifying cohesive subgraphs attract many attentions from different fields due to its numerous applications. Among many cohesive subgraph models, k-core is the most widely used model due to its simple and intuitive structure. In this paper, we formulate (p,n)-core in signed networks by extending k-core. (p,n)-core simultaneously guarantees sufficient internal positive edges and deficient internal negative edges. We formally prove that finding an exact (p,n)-core is NP-hard. Hence, we propose three efficient and effective algorithms to find a solution. Using real-world and synthetic networks, we demonstrate the superiority of our proposed algorithms.

Cohesive subgraph discovery in a network is one of the fundamental problems and investigated for several decades. In this paper, we propose the Overlapping Cohesive Subgraphs with Minimum degree (OCSM) problem which combines three key concepts for OCSM : (i) edge-based overlapping, (ii) the minimum degree constraint, and (iii) the graph density. To the best of our knowledge, this is the first work to identify overlapping cohesive subgraphs with minimum degree by incorporating the graph density. Since the OCSM problem is NP-hard, we propose two algorithms: advanced peeling algorithm and seed-based expansion algorithm. Finally, we show the experimental study with real-world networks to demonstrate the effectiveness and efficiency of our proposed algorithms.

We present a method for learning dynamics of complex physical processes described by time-dependent nonlinear partial differential equations (PDEs). Our particular interest lies in extrapolating solutions in time beyond the range of temporal domain used in training. Our choice for a baseline method is physics-informed neural network (PINN) [Raissi et al., J. Comput. Phys., 378:686--707, 2019] because the method parameterizes not only the solutions but also the equations that describe the dynamics of physical processes. We demonstrate that PINN performs poorly on extrapolation tasks in many benchmark problems. To address this, we propose a novel method for better training PINN and demonstrate that our newly enhanced PINNs can accurately extrapolate solutions in time. Our method shows up to 72% smaller errors than existing methods in terms of the standard L2-norm metric., Comment: Accepted by AAAI 2021