Information-Theoretic Analysis of Input Strokes in Visual Object Cutout
Information-Theoretic Analysis of Input Strokes in Visual Object Cutout
复制标题
视觉对象剪切中输入笔画的信息论分析
DOI：
10.1109/tmm.2010.2064759
复制
发表时间：
2010-12
影响因子：
7.3
通讯作者：
Yan, Shuicheng
中科院分区：
文献类型：
--
作者：
Mu, Yadong;Zhou, Bingfeng;Yan, Shuicheng
语义对象裁剪是各种图像编辑系统的基本单元。在一个典型的场景中，用户需要提供几个笔画，这些笔画指示部分像素作为图像背景或对象。然而，大多数现有的方法都是被动的，在这个意义上接受输入笔划，而不检查与用户意图的一致性。在这里，我们认为，积极的策略可能会减少互动的负担。在进行任何真实的分割计算之前，该程序可以粗略估计每个图像元素的不确定性，并主动为用户提供有用的建议。这样的预处理对于不知道用最佳笔划来馈送底层剪切算法的初学者特别有用。我们开发了这样一个活跃的对象剪切算法，名为ActiveCut，这使得它可以自动检测模糊给定当前用户提供的笔画，并合成“提示性笔画”作为反馈。一般来说，提示性笔画来自模糊的图像部分，具有最大的潜力，以减少标签的不确定性。用户可以在这些暗示性的笔画之后不断地改进他们的输入。以这种方式，用户-程序交互迭代的数量因此可以大大减少。具体地说，不确定性是由用户笔画和未标记的图像区域之间的互信息建模。为了确保ActiveCut以用户交互速率工作，我们采用基于超像素网格的图像表示，其计算取决于场景复杂度而不是原始图像分辨率。此外，它保留了2-D格的拓扑结构，因此更适合于并行计算。对于概率熵的计算，采用变分近似法进行加速。最后，基于子模函数理论，对所提贪婪算法的性能下界进行了理论分析。在MSRC图像数据集上进行了各种用户研究，以验证我们所提出的算法的有效性。
Semantic object cutout serves as a basic unit in various image editing systems. In a typical scenario, users are required to provide several strokes which indicate part of the pixels as image background or objects. However, most existing approaches are passive in the sense of accepting input strokes without checking the consistence with user's intention. Here we argue that an active strategy may potentially reduce the interaction burden. Before any real calculation for segmentation, the program can roughly estimate the uncertainty for each image element and actively provide useful suggestions to users. Such a pre-processing is particularly useful for beginners unaware of feeding the underlying cutout algorithms with optimal strokes. We develop such an active object cutout algorithm, named ActiveCut, which makes it possible to automatically detect ambiguity given current user-supplied strokes, and synthesize "suggestive strokes" as feedbacks. Generally, suggestive strokes come from the ambiguous image parts and have the maximal potentials to reduce label uncertainty. Users can continuously refine their inputs following these suggestive strokes. In this way, the number of user-program interaction iterations can thus be greatly reduced. Specifically, the uncertainty is modeled by mutual information between user strokes and unlabeled image regions. To ensure that ActiveCut works at a user-interactive rate, we adopt superpixel lattice based image representation, whose computation depends on scene complexity rather than original image resolution. Moreover, it retains the 2-D-lattice topology and is thus more suitable for parallel computing. While for the most time-consuming calculation of probabilistic entropy, variational approximation is utilized for acceleration. Finally, based on submodular function theory, we provide a theoretic analysis for the performance lower bound of the proposed greedy algorithm. Various user studies are conducted on the MSRC image dataset to validate the effectiveness of our proposed algorithm.
登录
查看更多内容
DOI：
10.1109/cvpr.2007.383498
发表时间：
2007-06
期刊：
2007 IEEE Conference on Computer Vision and Pattern Recognition
影响因子：
--
作者：
V. Sharma;James W. Davis
通讯作者：
V. Sharma;James W. Davis
影响因子：
3.7
作者：
T. Burr
通讯作者：
T. Burr
DOI：
10.1109/icmla.2007.54
发表时间：
2007-12
期刊：
Sixth International Conference on Machine Learning and Applications (ICMLA 2007)
影响因子：
--
作者：
Fei Wang;Xin Wang;Ta-Hsin Li
通讯作者：
Fei Wang;Xin Wang;Ta-Hsin Li
DOI：
--
发表时间：
2007-07
期刊：
--
影响因子：
--
作者：
Andreas Krause;Carlos Guestrin
通讯作者：
Andreas Krause;Carlos Guestrin
DOI：
10.1109/cvpr.2006.91
发表时间：
2006-06
期刊：
2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06)
影响因子：
--
作者：
C. Rother;T. Minka;A. Blake;V. Kolmogorov
通讯作者：
C. Rother;T. Minka;A. Blake;V. Kolmogorov