We present a dual adaptive garment slice adjustment technique for automatic resizing of apparel products with variant body shapes, and this can quickly generate clothed characters. Our first contribution is to propose novel fit evaluation method. When 2D pattern 3D draped have the same triangle topology connection, we calculate shape variable each output heat map simultaneously. For sewing newly targeted human body, fully method that conforms structure composed two stages. In coarse auto...

This paper introduces FSD-YOLOv8, an advanced YOLOv8 algorithm specifically adapted for the precise detection of flames and smoke in various complex environments. By integrating a preliminary image dehazing phase, our significantly enhances accuracy by reducing atmospheric confusion. Furthermore, we introduce application dilated convolutions within framework, which aids detecting intricate features associated with fire smoke. We present comprehensive evaluations approach using large,...

A fast segmentation algorithm of single medical image and sequence images based on active contour model are proposed in this paper. We give out the initial forecast 3D first, then, numerical solution is presented, send data blocks to different processors do calculations, finally combine results. At same time, parallel designing methods node allocation, task allocation load balancing researched. The experiment results show that method can significantly improve speed image.

Surface flattening plays an important role in the whole process of garment design. We proposed a novel method by using three‐dimensional triangle mesh this study. First, is flattened to two‐dimensional plane approximate original surface. The initial results are then used as preliminary guesses for subsequent optimizations. Considering that deformation energy real woven fabric related tensile or shear deformation, simplified model based on update distribution determine best pattern. An...

We present a novel virtual try‐on solution for fitting evaluation and pattern modification to design various types of garment speed up the process. In phase fit evaluation, we propose method producing two‐dimensional (2D) color maps by comparing 2D triangle mesh panel 3D garment, which can display outcome in real time. modification, prior condition based on maximum entropy coordinates more comprehensive mean value interpolation algorithm are proposed. By combination two deformation methods,...

Stereoscopic display vision is significantly impacted by the color distortion of left and right eye images. When human receives a specific range dissimilar information separately, visual system combines them into single through binocular fusion. In this study, we present experimental findings which compare accuracy common color-fusion model that was trained utilizing both linear fitting back-propagation neural networks. Patient contrast test data collected care professionals working in...

Global navigation satellite system (GNSS) has been widely used in various fields with potential security risks GNSS applications. We briefly describe the causes of vulnerability and introduce typical deception attacks. also point out that spoofing interference is more harmful. The background significance anti-spoofing technology are stated to classify summarize means measures recent years. Finally, advantages disadvantages current summarized discuss future research direction technology.

This paper presents a multiscale feature-guided residual convolutional neural network (ResCNN) method for synthetic aperture radar (SAR) image change detection. The proposed uses four kernels of different sizes to extract multi-scale features and fuses the make representation more powerful. In addition, introduces three blocks form CNN with than ten layers. deep residuals can ensure that model robust high-level semantic from avoid gradient dispersion problem in backpropagation. pseudo labels...

Displaying a variety of fabrics on customized character could help customers choose which fabric is more suitable for themselves and clothing. However, it not an easy task to show realistic garment virtual character. As result, we propose stable finite element method (FEM) model approximate stretching behaviors. At first, measure four kinds cloth materials with measurement techniques research elastic deformations in real samples. Then, use the parameter optimization by fitting data. For...

<abstract><p>Computing the minimal axiom sets (MinAs) for an unsatisfiable class is important task in incoherent ontology debugging. Ddebugging ontologies based on patterns (DOBP) a pattern-based debugging method that uses set of heuristic strategies four patterns. Each pattern represented as directed graph and depth-first search strategy used to find paths relevant MinAs class. However, DOBP inefficient when large with lot classes. To solve problem, we first extract module...

Gossip learning is one of the most advanced approaches in decentralized techniques and performs almost as well centralized training. However, gossip inherently vulnerable to malicious attacks from adversarial nodes due its flat architecture, which leads performance degradation collaborative training nodes. To reduce negative impact nodes, we propose an effective strategy protect distributed process. The results show that proposed robust attacks.

With the maturity of 3D and AR technologies, many e-commerce platforms have launched function virtual fitting rooms. Although it attracted a lot attention, real experience is not good. At present, main problems are inaccurate in clothing version size, dressing effect quite different. Therefore, platform to customize personalized has potential application value. In this study, we consider trying on different detailed mannequin increase experience. First, human body database established, then...

The uncertainty and randomness of large-scale photovoltaic (PV) generation will seriously affect the safety economy power systems. In order to mitigate negative impact grid-connected solar PV, it is important accurately predict PV power. view this, this paper first proposes a novel hybrid method based on non-pooling convolutional neural network deep deterministic policy gradient (DDPG) for ultra-short-term forecasting. DDPG model used implement feature training task, NPCNN introduced into...