动作识别与3D剩余注意力和交叉
1School of Computer Science and Technology, Beijing Jiaotong University, Beijing 100044, China.
Entropy (Basel, Switzerland)
|April 26, 2025
概括
本研究引入了用于人类活动识别的3D残留注意网络 (3DRFNet). 这种新型网络有效地捕捉了时空特征,在基准数据集上实现了最先进的准确性.
科学领域:
- 计算机科学 计算机科学
- 人工智能的人工智能
- 机器学习 机器学习
背景情况:
- 人类活动识别 (HAR) 对于监控和人机交互等应用至关重要.
- 现有的方法往往难以有效地捕捉视频中的复杂的时空动态.
研究的目的:
- 提出一种新的深度学习模型,即3D残留注意网络 (3DRFNet),用于增强人类活动识别.
- 提高模型从视频数据中学习歧视性时空特征的能力.
主要方法:
- 开发了一个3D残留注意网络 (3DRFNet),将通道和空间注意机制集成到3D ResNet框架内.
- 整合了快速里埃卷积 (FFC) 来增强时间和空间特征提取.
- 用于模型训练和反向传播的利用交叉损失.
主要成果:
- 3DRFNet在人类行动识别任务上取得了最先进的 (SOTA) 性能.
- 在HMDB-51数据集上达到91.7%的高精度,在UCF-101数据集上达到98.7%.
- 证明了卓越的识别准确性和稳定性,特别是在捕捉关键的行为特征.
结论:
- 拟议的3DRFNet有效地学习了人类活动识别的时空表征.
- 注意力机制和FFC显著有助于模型的增强性能.
- 3DRFNet为基于视频的人类行为识别提供了强大而准确的解决方案.
相关概念视频
Three-Dimensional Force System:Problem Solving
A three-dimensional force system refers to a scenario in which three forces act simultaneously in three different directions. This type of problem is commonly encountered in physics and engineering, where it is necessary to calculate the resultant force on the system, which can then be used to predict or analyze the behavior of the object or structure under consideration.
To solve a three-dimensional force system, first resolve each force into its respective scalar components. Do this using...
To solve a three-dimensional force system, first resolve each force into its respective scalar components. Do this using...
Relative Motion Analysis using Rotating Axes
Consider a component AB undergoing a linear motion. Along with a linear motion, point B also rotates around point A. To comprehend this complex movement, position vectors for both points A and B are established using a stationary reference frame.
However, to express the relative position of point B relative to point A, an additional frame of reference, denoted as x'y', is necessary. This additional frame not only translates but also rotates relative to the fixed frame, making it instrumental in...
However, to express the relative position of point B relative to point A, an additional frame of reference, denoted as x'y', is necessary. This additional frame not only translates but also rotates relative to the fixed frame, making it instrumental in...
Relative Motion Analysis using Rotating Axes-Problem Solving
Consider a crane whose telescopic boom rotates with an angular velocity of 0.04 rad/s and angular acceleration of 0.02 rad/s2. Along with the rotation, the boom also extends linearly with a uniform speed of 5 m/s. The extension of the boom is measured at point D, which is measured with respect to the fixed point C on the other end of the boom. For the given instant, the distance between points C and D is 60 meters.
Here, in order to determine the magnitude of velocity and acceleration for point...
Here, in order to determine the magnitude of velocity and acceleration for point...
Depth Perception and Spatial Vision
Depth perception is the ability to perceive objects three-dimensionally. It relies on two types of cues: binocular and monocular. Binocular cues depend on the combination of images from both eyes and how the eyes work together. Since the eyes are in slightly different positions, each eye captures a slightly different image. This disparity between images, known as binocular disparity, helps the brain interpret depth. When the brain compares these images, it determines the distance to an object.


