在CNN架构中增强色彩选择性的长跳转连接
Oscar Sanchez-Cesteros1, Mariano Rincon1, Margarita Bachiller1
1Department of Artificial Intelligence, National University of Distance Education (UNED), 28040 Madrid, Spain.
Sensors (Basel, Switzerland)
|September 9, 2023
概括
将长跳转连接 (LSC) 添加到卷积神经网络 (CNN) 中,可以提高图像分类准确性和颜色选择性. 这种修改允许更深层的网络层访问早期层的信息,增强自然场景的特征提取.
科学领域:
- 计算机视觉 计算机视觉
- 神经科学启发的人工智能
- 机器学习 机器学习
背景情况:
- 卷积神经网络 (CNN) 在自然图像数据集中表现出有限的颜色选择性.
- CNNs的灵感来源于视觉皮层的等级结构.
- 现有的CNN架构可能无法充分利用更深层的低级视觉信息.
研究的目的:
- 为了提高CNN的颜色选择性和分类准确性.
- 研究长跳连接 (LSC) 对CNN性能的影响.
- 改善CNN中低级别和高级别特征的整合.
主要方法:
- 建议在CNN特征提取阶段的第一个和最后一个块之间添加一个长跳过连接 (LSC).
- 将LSC策略应用于已建立的CNN架构.
- 定量和质量分析了准确性和颜色选择性方面的改进.
- 开发了一种用于特征地图可视化的新色彩表示方法.
主要成果:
- 一般来说,LSCs可以提高CNN分类的准确性.
- 拟议的LSC在标准跳过连接上表现出优越的性能.
- 在加上LSCs的CNN中观察到增强的颜色选择性.
- 新的颜色表示方便了可管理的特征图的定性分析.
结论:
- 长跳过连接有效地提高了CNN的准确性和颜色选择性.
- 通过LSC集成浅层和深层信息,有利于处理自然场景.
- 拟议的LSC战略为开发更感知一致的AI视觉系统提供了一个有希望的方向.
相关概念视频
Color Vision
611
Color perception begins in the retina, the light-sensitive layer at the back of the eye. Two main theories explain how colors are seen: the trichromatic theory and the opponent-process theory. The trichromatic theory, proposed by Thomas Young in 1802 and extended by Hermann von Helmholtz in 1852, suggests that color vision is based on three types of cone receptors in the retina. These cones are sensitive to different but overlapping ranges of wavelengths corresponding to red, blue, and green.
611
Reducing Line Loss
173
In a three-phase circuit, line loss is an indicator of energy dissipated as heat due to the resistance of transmission lines. To address this, incorporating transformers into the system—a step-up transformer at the source and a step-down transformer at the load—is a strategic solution. Two three-phase transformers are introduced to improve this.
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss...
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss...
173


