A novel broadband bandpass Frequency Selective Surface (FSS) is designed with the intention of surpassing the performance of previous designs. Through the utilization of cascading multilayers of capacitive patches and...
详细信息
Class-incremental learning (CIL) aims to train a model to learn new classes from non-stationary data streams without forgetting old ones. In this paper, we propose a new kind of connectionist model by tailoring neural...
详细信息
Class-incremental learning (CIL) aims to train a model to learn new classes from non-stationary data streams without forgetting old ones. In this paper, we propose a new kind of connectionist model by tailoring neural unit dynamics that adapt the behavior of neural networks for CIL. In each training session, it introduces a supervisory mechanism to guide network expansion whose growth size is compactly commensurate with the intrinsic complexity of a newly arriving task. This constructs a near-minimal network while allowing the model to expand its capacity when cannot sufficiently hold new classes. At inference time, it automatically reactivates the required neural units to retrieve knowledge and leaves the remaining inactivated to prevent interference. We name our model AutoActivator, which is effective and scalable. To gain insights into the neural unit dynamics, we theoretically analyze the model's convergence property via a universal approximation theorem on learning sequential mappings, which is under-explored in the CIL community. Experiments show that our method achieves strong CIL performance in rehearsal-free and minimal-expansion settings with different backbones. Copyright 2024 by the author(s)
Using an Iterated Dilated Convolutional Neural Network (IDCNN) in Chinese Named Entity Recognition (NER) helps capture local information. However, the contribution of information from different positions in a sentence...
详细信息
Active object detection (AOD) aims to guide a robot to make appropriate moving actions to get close to the target object, which is significant for the service robot to complete tasks in the indoor household environmen...
详细信息
This paper introduces a quadratic programming-based stable motion control method for humanoid robots. This method is based on the virtual model and the optimal force distribution. We added a six-dimensional mixed cons...
详细信息
Autonomous driving systems open up a new frontier in the automotive industry, offering new possibilities for future transportation with increased efficiency and a comfortable experience. However, object detection in a...
详细信息
Since Unmanned Aerial Vehicle (UAV) swarm emerged with the merit of high efficiency, flexibility and robustness in search tasks, related research has sprung up in recent years. However, unknown target locations make s...
详细信息
The power quality is an important index of power industry. It is important to identify the type of power quality disturbances (PQDs) accurately. To address the low precision of PQDs feature extraction and the complexi...
详细信息
NO and NO2 are important gases that can lead to the formation of photochemical smog and acid rain. Therefore, it is important to develop high-performance nitrogen oxide sensors. Here, the electronic structure, spin tr...
详细信息
With the rapid development of Cyber-Physical Systems (CPS), security concerns have increasingly garnered attention. This paper focuses on the security threats faced by smart grid systems, particularly false data injec...
详细信息
暂无评论