arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1706.05358cs.LGstat.ML
收藏

基于自适应Siamese网络的局部特征描述子学习

Local Feature Descriptor Learning with Adaptive Siamese Network

University of California, Santa Barbara(加州大学圣塔芭芭拉分校);Fuji Xerox Palo Alto Laboratory(富士施乐帕洛阿尔托实验室)

Chong Huang, Qiong Liu, Yan-Ying Chen, Kwang-Ting, Cheng

摘要:

虽然最近在深度神经网络中的进展已经导致了可学习的局部特征描述符的发展,但是对于神经网络的必要大小的估计没有明确的答案。具体而言,局部特征在低维空间中表示,因此神经网络应该具有更紧凑的结构。局部特征描述符学习所需的小网络可能对初始条件和学习参数敏感,并且更有可能陷入局部极小值。为了解决上述问题,我们引入了一种基于神经元激活的自适应修剪Siamese架构来学习局部特征描述符,使网络在更复杂的网络中具有更高的计算效率和更高的识别率。我们的实验表明,我们学习的局部特征描述符优于国家的最先进的方法在补丁匹配。

英文摘要:

Although the recent progress in the deep neural network has led to the development of learnable local feature descriptors, there is no explicit answer for estimation of the necessary size of a neural network. Specifically, the local feature is represented in a low dimensional space, so the neural network should have more compact structure. The small networks required for local feature descriptor learning may be sensitive to initial conditions and learning parameters and more likely to become trapped in local minima. In order to address the above problem, we introduce an adaptive pruning Siamese Architecture based on neuron activation to learn local feature descriptors, making the network more computationally efficient with an improved recognition rate over more complex networks. Our experiments demonstrate that our learned local feature descriptors outperform the state-of-art methods in patch matching.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑