In this work, we present MUSE (Model-based Uncertainty-aware Similarity Estimation), a training-free framework for model-based zero-shot 2D object detection and segmentation. First, MUSE incorporates 2D multi-view templates from 3D unseen objects and 2D object proposals from the input query image, respectively. In the embedding stage, we propose a new feature embedding scheme which integrates class and patch embeddings. Specifically, the patch embeddings are normalized using the generalized mean pooling (GeM). In the matching stage, a joint similarity score is introduced, which integrates an absolute score and a relative score. Finally, we update the similarity score using an uncertainty-aware object prior. MUSE achieves state-of-the-art performance on the BOP Challenge 2025, ranking first in the Classic Core, H3, and Industrial tracks—without any additional training or fine-tuning. Therefore, we believe that MUSE is a promising framework for zero-shot 2D object detection and segmentation.

This study proposes a novel approach that quantifies the local distribution of point clouds to define Dimensionality, incorporates it into the existing 3D coordinate system to construct a 4D coordinate system, and employs Hilbert Curve to efficiently improve registration speed. By analyzing the distributional characteristics of point clouds using Hellinger Distance and introducing Hilbert Curve-based nearest neighbor search, the proposed method reduces computational overhead while enhancing registration accuracy. The algorithm was evaluated on the Kitti and Stanford Bunny datasets [1], demonstrating lower performance compared to existing algorithms in the simple urban environment of the Kitti dataset but achieving outstanding registration performance in the complex structure of the Stanford Bunny dataset, even in cases involving partial overlaps and outliers. In particular, the combination of Dimensionality and Hilbert Curve-based search significantly improved registration quality and stability in complex datasets. This approach can be extended to LIO-based odometry systems or utilized in Color-based ICP applications.

In this paper, we propose a novel densely-packed object detection method based on advanced weighted Hausdorff distance (AWHD) and hard negative-aware anchor (HNAA) attention. Densely-packed object detection is more challenging than conventional object detection due to the high object density and small-size objects. To overcome these challenges, the proposed AWHD improves the conventional weighted Hausdorff distance and obtains an accurate center area map. Using the precise center area map, the proposed HNAA attention determines the relative importance of each anchor and imposes a penalty on hard negative anchors. Experimental results demonstrate that our proposed method based on the AWHD and HNAA attention produces accurate densely-packed object detection results and comparably outperforms other state-of-the-art detection methods. The code is available at ${\color{Blue} \text{here}}$.

In this paper, we propose a novel network pruning method using the proposed relative depth-wise separable convolutions and softmax attention channel pruning. The relative depthwise separable convolution enhances conventional depth-wise separable convolutions by enabling the channel interaction, which can prevent accuracy drops even after severe pruning. The softmax attention channel pruning probabilistically expresses the importance of filters and removes unimportant channels efficiently. Experimental results demonstrate that our pruning method outperforms other state-of-the-art pruning methods in terms of Flops, parameters, and top-1 classification accuracy.

We present a novel method for the upright adjustment of 360° images. Our network consists of two modules, which are a convolutional neural network (CNN) and a graph convolutional network (GCN). The input 360° images is processed with the CNN for visual feature extraction, and the extracted feature map is converted into a graph that finds a spherical representation of the input. We also introduce a novel loss function to address the issue of discrete probability distributions defined on the surface of a sphere. Experimental results demonstrate that our method outperforms fully connected-based methods.

We present a novel object detection method called the multi-domain attentive detection network (MDADN). For robust object detection, we do not only use red, green, and blue (RGB) image data but also infrared data. The MDADN adds attention modules to each layer to weigh multi-domains of data differently in a channel-wise and spatial-wise manner, which yields channel- and spatial-aware networks. The MDADN accurately detects objects (e.g., vehicles) in challenging environments, including nighttime, rainy and foggy days, and environments where even humans find it difficult to detect objects. Experimental results on the FLIR dataset demonstrate that the MDADN outperforms state-of-the-art methods in terms of accuracy and speed. The ablation study shows that the attention module and the fusion of multiple data sources help to considerably improve the accuracy.

Magnetic resonance electrical property tomography (MREPT) is a technique used to extract the electrical properties of tissues (conductivity in particular) using a magnetic resonance imaging system. In this paper, we propose an improved data acquisition scheme for the electrical property tomography technique by utilizing T2modulation in fast spin echo (FSE) imaging. This technique was motivated by a numerical analysis of conductivity reconstruction in the frequency domain; results reveal the spatial frequency-dependent noise texture of conventional methods. A data-acquisition scheme using the FSE sequence was formulated to concentrate the signal within a specific frequency range where notable noise amplification is observed in the conventional method. Through numerical studies, the performance of the proposed acquisition was investigated. Furthermore, a compensation scheme was applied to reduce quantification errors due to tissue-specific T2modulation, which is inherent in FSE imaging. The technique was applied to phantom and in vivo experiments. Results showed improved conductivity contrasts in both experiments, as compared with conventional MREPT methods.

As the popularity of mobile video streaming services has been increasing rapidly, mobile network operators are faced with big challenges how to guarantee the quality of video streaming services especially in the wireless mobile network. In this paper, we were focused on the real-time live streaming services in LTE network which has fluctuated radio environment that requires special treatments for better QoS regarding delicate properties, such as limited buffered data and wide range of data throughput (i.e. up to 150Mbps). In the main part of this paper, we will provide experimental results of field test with different network environment over Mobile MMT(Mpeg Media Transport)[1] protocol which is implemented in Oksusu (i.e. SK mobile IPTV service) over LTE network.

An algorithm for technology mapping of combinational and sequential logic networks is proposed and applied to mapping into K-input lookup-tables (K-LUTs). The new algorithm avoids the hurdle of computing all K-input cuts while preserving the quality of the results, in terms of area and depth. The memory and runtime of the proposed algorithm are linear in circuit size and quite affordable even for large industrial designs. For example, computing a good quality 6-LUT mapping of an AIG with 1M nodes takes 150Mb of RAM and 1 minute on a typical laptop. An extension of the algorithm allows for sequential mapping, which searches the combined space of all possible mappings and retimings. This leads to an 18–22% improvement in depth with a 3–5% LUT count penalty, compared to combinational mapping followed by retiming.