摘要
为实现用较少的训练样本对高分辨距离像进行识别,文中提出一种采用多任务稀疏学习的统计建模方法.该方法将各帧训练样本的统计建模视为单一的任务,由于各帧训练样本间不是完全独立而是相互关联的,因此,设定所有帧的训练样本采用同一个字典以实现帧间信息的共享.由于目标的不同以及同一目标的方位敏感性,通常很难确定各训练帧的相关性,而不相关任务间的联合学习将会降低识别性能.因此,采用Bernoulli-Beta先验根据给定训练数据自动学出每一帧需要的原子,而通过不同帧间共享的原子个数就可以判断它们的相关性,从而实现自适应的多任务学习.基于实测高分辨距离像数据的识别实验,证明了文中方法的有效性.
A statistical modeling method based on multitask sparse learning is proposed to realize the recognition of the high resolution range profile（HRRP）with a small training data size.The statistical modeling of each training aspect-frame is considered as a single task in our method.Since the training aspect-frames are not independent but inter-related,they can share a compact dictionary to make full use of the information.However,with the different targets and the aspect sensitivity of the same target,it is usually hard to assess the task relatedness,and joint learning with unrelated tasks may degrade the recognition performance.Therefore,we adopt the Bernoulli-Beta prior to learn the needed atoms of each aspect-frame automatically with the given training data.Then the relatedness between frames is determined by the number of shared atoms,and multitask learning can be realized adaptively.The recognition experiments of the measured HRRP data demonstrate the performance of the proposed method.
出处
《西安电子科技大学学报》
EI
CAS
CSCD
北大核心
2016年第2期23-28,共6页
Journal of Xidian University
基金
国家自然科学基金资助项目(61271024
61201296
61322103)
全国优秀博士学位论文作者专项资金资助项目(FANEDD-201156)
中央高校基本科研业务费专项资金资助项目(K5051302010)
