Journals
  Publication Years
  Keywords
Search within results Open Search
Please wait a minute...
For Selected: Toggle Thumbnails
High-Level Semantic Distillation for Incremental and Continuous Object Detection
Mengxue KANG, Jinpeng ZHANG, Zhe MA, Xuhui HUANG, Yating LIU, Zizhuang SONG
Modern Defense Technology    2024, 52 (1): 41-48.   DOI: 10.3969/j.issn.1009-086x.2024.01.006
Abstract48)   HTML3)    PDF (624KB)(51)       Save

Modern defence requires intelligent perception algorithms to possess incremental and continuous learning capabilities in complex open scenarios, while traditional deep learning methods are based on closed training with the entire dataset, which limits their application ability and usage scope. Existing continuous learning algorithms face the problem of catastrophic knowledge forgetting. This paper proposes for the first time an incremental continuous target detection method based on the distillation of high-level semantic features, which guides the selection of high-value underlying features through high-level semantic features and distills the feature from the teacher model to the student model, thus effectively facilitating the transfer of knowledge of the old task and alleviating catastrophic knowledge forgetting. Experiments on the public image dataset MS COCO show that this method outperforms the previous best method for target detection in all types of continuous learning scenarios, which is expected to promote the generation of continuous learning capability and autonomous attempts of intelligent systems in open-world setting.

Table and Figures | Reference | Related Articles | Metrics
Multi-ballistic Trajectory Planning and Simulation Based on Multi-objective Optimization Algorithm
Yaohua ZHANG, Yaning XU, Shaojie GAO, Peng ZHANG, Renti LIU, Xiaolong SHI, Xi YANG
Modern Defense Technology    2023, 51 (5): 77-85.   DOI: 10.3969/j.issn.1009-086x.2023.05.010
Abstract52)   HTML5)    PDF (1737KB)(133)       Save

Multi-missile coordination is a hot research direction of missile weapons with high research value, and how to plan the flight trajectory of multiple missiles is an important research content. In this paper, a multi-ballistic trajectory simulation method based on a multi-objective evolutionary algorithm based on decomposition (MOEA/D) algorithm is proposed, which takes the glide phase of boost glide missile as the research object and aims at planning multiple high-quality trajectories. The maximum range and the maximum final velocity are selected as the objective functions to realize the trajectory planning of the glide phase. Meanwhile, in order to reduce the impact of the initial population on the multi-objective optimization problem, the pseudo-spectral method is used to provide high-quality initial populations for the multi-objective optimization algorithm. The simulation results show that the method can obtain multi-ballistic trajectories with optimal range and final velocity.

Table and Figures | Reference | Related Articles | Metrics