所以我们又给出一种自适应修正自然梯度算法，用一个合适的距离测度函数来控制步长和动力因子。
So we present an adaptive improved natural gradient algorithm, which use an appropriate estimation function to control the step-size and the momentum factor.
另一方面，对线性融合后的特征，利用距离测度学习算法从图像训练集中学到一个距离函数以更恰当的反映图像内容间的相似度。
On the other hand, a proper distance metric, which reflects the semantic similarity between images, is obtained by using a distance metric learning algorithm.
应用推荐