GDVIFNet:一个生成的深度和可见图像融合网络,具有边缘特征指导,用于突出物体检测
Xiaogang Song1, Yuping Tan2, Xiaochang Li3
1Xi'an University of Technology, School of Computer Science and Engineering, Xi'an, 710048, China; Engineering Research Center of Human-machine integration intelligent robot, Universities of Shaanxi Province, Xi'an, 710048, China.
概括
本研究介绍了GDVIFNet,这是一种新的突出物体检测 (SOD) 方法,仅使用RGB图像. 它有效地生成深度信息,并通过边缘功能来改进它,以提高复杂环境中的性能.
科学领域:
- 计算机视觉 计算机视觉
- 人工智能的人工智能
- 图像处理 图像处理
背景情况:
- 在复杂的环境中,突出物体检测 (SOD) 的性能不足于最佳.
- 现有的SOD方法通常需要额外的深度或热传感器,增加成本和不便.
研究的目的:
- 提出GDVIFNet,一种新的方法,用于仅使用单个RGB图像检测突出物体.
- 通过利用生成的深度信息来克服专门硬件的局限性.
主要方法:
- 使用深度任何东西从RGB输入生成深度图像.
- 采用自主监督技术,从生成的深度图像中生成噪声和文物清除的边缘特征.
- 使用双分支架构 (CNN和变压器) 与新型融合模块 (STIU,SCF,CETSF) 进行功能集成.
主要成果:
- 在多个突出物体检测数据集上,GDVIFNet实现了最先进的性能.
- 拟议的方法有效地处理噪声和人工物在生成的深度数据.
- 证明了融合RGB,生成深度和边缘特征的有效性.
结论:
- GDVIFNet提供了一种具有成本效益和方便的解决方案,用于突出物体检测.
- 生成的深度和自我监督的边缘特征的集成显著提高了SOD性能.
- 具有先进的融合机制的双分支架构对于复杂的视觉场景证明是有效的.
相关概念视频
Depth Perception and Spatial Vision
483
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.
483
Association Areas of the Cortex
4.6K
Association areas are regions of the cerebral cortex that do not have a specific sensory or motor function. Instead, they integrate and interpret information from various sources to enable higher cognitive processes such as memory, learning, and decision-making. Some key association areas include the following:
Prefrontal Association Area: This area is located in the frontal lobe and is involved in planning, decision-making, and moderating social behavior. It connects with primary motor areas,...
Prefrontal Association Area: This area is located in the frontal lobe and is involved in planning, decision-making, and moderating social behavior. It connects with primary motor areas,...
4.6K
Vision
52.4K
Vision is the result of light being detected and transduced into neural signals by the retina of the eye. This information is then further analyzed and interpreted by the brain. First, light enters the front of the eye and is focused by the cornea and lens onto the retina—a thin sheet of neural tissue lining the back of the eye. Because of refraction through the convex lens of the eye, images are projected onto the retina upside-down and reversed.
52.4K


