通过利用不确定性意识的神经网络,实现精确的单一图像沙尘除尘
Bingcai Wei1, Hui Liu1, Chuang Qian2
1School of Computer Science, Wuhan University, Wuhan, China.
Frontiers in neurorobotics
|September 26, 2025
概括
本研究介绍了层次互动不确定性意识网络 (HIUNet),用于有效的单一图像去除沙尘. HIUNet使用贝叶斯神经网络来解决环境不确定性,并使用特征频率选择来恢复高质量的图像.
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:
- 深度学习在图像除尘方面表现有前途,但与异质的环境不确定性作斗争.
- 现有的方法往往无法充分解决沙子和尘埃造成的复杂降解.
研究的目的:
- 开发一种新的深度学习框架,分层交互式不确定性意识网络 (HIUNet),用于强大的单一图像沙尘除尘.
- 为了有效地建模和减轻尘埃环境中固有的不确定性,以改善图像恢复.
主要方法:
- 利用贝叶斯神经网络进行强大的浅层特征提取.
- 使用预先训练有素的编码器和轻量级解码器进行初始图像重建.
- 实施了特征频率选择机制,以识别和保留有价值的特征,同时抑制噪音.
- 集成了一个功能增强模块来完善初步恢复.
主要成果:
- HIUNet在从退化输入中重建高质量的清洁图像方面表现出卓越的性能.
- 在Sand11K数据集上的实验验验证了该方法在各种降解水平的有效性.
- 该框架成功模拟了不确定性,并选择了重建的突出特征.
结论:
- 通过解决环境不确定性,HIUNet提供了一种有效的解决方案,用于单一图像去除沙尘.
- 贝叶斯网络和特征频率选择的结合是高质量的图像重建的关键.
- 未来的工作重点是将框架扩展到极端沙子场景.
相关概念视频
00:59Uncertainty: Overview
1.6K
In analytical chemistry, we often perform repetitive measurements to detect and minimize inaccuracies caused by both determinate and indeterminate errors. Despite the cares we take, the presence of random errors means that repeated measurements almost never have exactly the same magnitude. The collective difference between these measurements - observed values - and the estimated or expected value is called uncertainty. Uncertainty is conventionally written after the estimated or expected value.
1.6K
00:59Propagation of Uncertainty from Random Error
1.8K
An experiment often consists of more than a single step. In this case, measurements at each step give rise to uncertainty. Because the measurements occur in successive steps, the uncertainty in one step necessarily contributes to that in the subsequent step. As we perform statistical analysis on these types of experiments, we must learn to account for the propagation of uncertainty from one step to the next. The propagation of uncertainty depends on the type of arithmetic operation performed on...
1.8K