This study investigates the evaluation of the representativeness of datasets in autonomous driving systems using statistical distance metrics. Eight different datasets, created in the Carla simulation environment, cov...
详细信息
Hybrid CPUs are a response from industry to the increasingly difficult problem of power management in desktop-class processors. Hybrid CPUs comprise a big/performance core for regular tasks and a small/efficiency core...
详细信息
Machine learning techniques, particularly fuzzy Support Vector Machines (fuzzy SVM), have demonstrated effectiveness in sentiment classification within social networks. However, the evolving complexity of social netwo...
详细信息
Crease design is crucial to the function and application scope of origami projects;however, for the Yoshimura origami pattern, the connection between origami two-dimensional (2D) creases and three-dimensional (3D) for...
详细信息
Crease design is crucial to the function and application scope of origami projects;however, for the Yoshimura origami pattern, the connection between origami two-dimensional (2D) creases and three-dimensional (3D) forms has not been established, which greatly limits the research and application of origami structures. In this paper, a crease design theory for the Yoshimura origami pattern is proposed to create 3D foldable origami structures. Afterward, the flat origami folding process is described, and the corresponding end trajectory equations are derived from a crease chain folding analysis performed on the element block and the expansion block. Next, the positional and angular constraint equations for single-variable, double-variable, and triple- variable foldable forms with only element blocks are systematically discussed. Finally, a detailed folding simulation analysis is performed based on the positional and angular constraint equations, and the foldable form family diagrams with the corresponding angular value intersection curve constraints are summarized. The theory is verified by finite element programs and experiments. This work provides an effective theory for the design of origami creases and new ideas and methods for the study of the Yoshimura pattern and other origami patterns.
In previous power system predictions, the accuracy of predicting extreme electricity consumption peaks was often insufficient. This study aims to explore the differences in power data prediction among three models: Bi...
详细信息
Modern automation schemes have substantially improved by using standard communication protocols, upgraded technologies and user-friendly tools in power distribution domain. Nevertheless, this fundamental functionality...
详细信息
Recent decades have seen remarkable progress in the field of robotic-assisted construction. Cable-driven parallel robots (CDPRs) emerge as promising tools for automating construction processes, due to their advantageo...
详细信息
Pre-trained language models (PLMs) are commonly used for various downstream natural language processing tasks via ***, recent studies have demonstrated that PLMs are vulnerable to backdoor attacks, which can mislabel ...
详细信息
Pre-trained language models (PLMs) are commonly used for various downstream natural language processing tasks via ***, recent studies have demonstrated that PLMs are vulnerable to backdoor attacks, which can mislabel poisoned samples to target outputs even after a vanilla fine-tuning *** key challenge for defending against the backdoored PLMs is that end users who adopt the PLMs for their downstream tasks usually do not have any knowledge about the attacking strategies, such as *** tackle this challenge, in this work, we propose a backdoor mitigation approach, PURE, via head pruning and normalization of attention *** idea is to prune the attention heads that are potentially affected by poisoned texts with only clean texts on hand and then further normalize the weights of remaining attention heads to mitigate the backdoor *** conduct experiments to defend against various backdoor attacks on the classification *** experimental results show the effectiveness of PURE in lowering the attack success rate without sacrificing the performance on clean *** code is available at https://***/xingyizhao/PURE. Copyright 2024 by the author(s)
Force-based algorithms for ab initio atomic structure relaxation, such as conjugate gradient methods, usually get stuck in the line minimization processes along search directions, where expensive ab initio calculation...
详细信息
Force-based algorithms for ab initio atomic structure relaxation, such as conjugate gradient methods, usually get stuck in the line minimization processes along search directions, where expensive ab initio calculations are triggered frequently to test trial positions before locating the next iterate. We present a force-based gradient descent method, WANBB, that circumvents the deficiency. At each iteration, WANBB enters the line minimization process with a trial step size capturing the local curvature of the energy surface. The exit is controlled by a nonrestrictive criterion that tends to accept early trials. These two ingredients streamline the line minimization process in WANBB. The numerical simulations on nearly 80 systems with good universality demonstrate the considerable compression of WANBB on the cost for the unaccepted trials compared with conjugate gradient methods. We also observe across the board significant and universal speedups as well as the superior robustness of WANBB over several widely used methods. The latter point is theoretically established. The implementation of WANBB is pretty simple, in that no a priori physical knowledge is required and only three parameters are present without tuning.
Despite significant advancements in remote sensing multimodal learning, particularly in image-image feature fusion, the exploration of audio-image feature fusion remains insufficient. Given the complexity and redundan...
详细信息
暂无评论