基于雷达的跨域人类行为识别使用物理信息增强和多源域对抗神经网络
IEEE journal of biomedical and health informatics
|August 27, 2025
概括
基于物理的数据增强 (PIDA) 增强了基于雷达的人类行为识别,特别是在老龄化人口中. 这种方法通过模拟现实的雷达条件和运动变化来提高域的概括性.
科学领域:
- 雷达信号处理
- 人类行为分析的机器学习
- 数据增强技术
背景情况:
- 跌倒是安全问题, 尤其是老年人.
- 基于雷达的人类行为识别,特别是落,已被广泛研究.
- 由于有限的标记数据,目前的方法普遍性较差.
研究的目的:
- 提出一个新的基于物理的数据增强 (PIDA) 战略.
- 通过使用FMCW雷达的微多普勒签名 (m-DS) 来增强人类行为识别的域泛化 (DG).
- 提高落检测系统的准确性和稳定性.
主要方法:
- 开发了PIDA,包括距离数据增强 (DDA) 和行为模式数据增强 (BPDA).
- DDA模拟了跨范围的电磁衰减;BPDA在保留动力学的同时多样化了运动风格.
- 集成PIDA与多源域对抗神经网络 (MSDAN) 进行转移学习和不变特征提取.
主要成果:
- 与基线,传统光学数据增强 (TODA) 和生成数据增强 (GDA) 相比,PIDA显著提高了平均GD准确性.
- 在自己的数据集中,PIDA比基线,TODA和GDA分别提高了1. 52%,2. 48%和4. 88%的精度.
- 在公共数据集上,PIDA相对于基线,TODA和GDA分别实现了3.67%,6.29%和9.78%的改进,其中GD差距很小.
结论:
- 在基于雷达的人类行为识别中, PIDA 有效地解决了数据短缺问题.
- 拟议的方法显示出用于落检测的优异域泛化能力.
- PIDA提供了一个有前途的方法,用于强大而准确的与跌倒有关的人类行为识别系统.
相关概念视频
02:39Predator-Prey Interactions
17.2K
Predators consume prey for energy. Predators that acquire prey and prey that avoid predation both increase their chances of survival and reproduction (i.e., fitness). Routine predator-prey interactions elicit mutual adaptations that improve predator offenses, such as claws, teeth, and speed, as well as prey defenses, including crypsis, aposematism, and mimicry. Thus, predator-prey interactions resemble an evolutionary arms race.
17.2K
02:26Defenses Against Pathogens and Herbivores
21.7K
Plants present a rich source of nutrients for many organisms, making it a target for herbivores and infectious agents. Plants, though lacking a proper immune system, have developed an array of constitutive and inducible defenses to fend off these attacks.
21.7K
01:05Collisions in Multiple Dimensions: Introduction
6.3K
It is far more common for collisions to occur in two dimensions; that is, the initial velocity vectors are neither parallel nor antiparallel to each other. Let's see what complications arise from this. The first idea is that momentum is a vector. Like all vectors, it can be expressed as a sum of perpendicular components (usually, though not always, an x-component and a y-component, and a z-component if necessary). Thus, when the statement of conservation of momentum is written for a...
6.3K
01:06Collisions in Multiple Dimensions: Problem Solving
4.5K
In multiple dimensions, the conservation of momentum applies in each direction independently. Hence, to solve collisions in multiple dimensions, we should write down the momentum conservation in each direction separately. To help understand collisions in multiple dimensions, consider an example.
A small car of mass 1,200 kg traveling east at 60 km/h collides at an intersection with a truck of mass 3,000 kg traveling due north at 40 km/h. The two vehicles are locked together. What is the...
A small car of mass 1,200 kg traveling east at 60 km/h collides at an intersection with a truck of mass 3,000 kg traveling due north at 40 km/h. The two vehicles are locked together. What is the...
4.5K