小样本困境下的深度学习图像识别综述

图像识别是图像研究领域的核心问题,解决图像识别问题对人脸识别、自动驾驶、机器人等各领域研究都有重要意义.目前广泛使用的基于深度神经网络的机器学习方法,已经在鸟类分类、人脸识别、日常物品分类等图像识别数据集上达到了超过人类的水平,同时越来越多的工业界应用开始考虑基于深度神经网络的方法,以完成一系列图像识别业务.但是深度学...

Survey on Deep Learning Image Recognition in Dilemma of Small Samples

Present machine learning methods have reached a higher level than human intelligence in image recognition and other tasks. However, recent machine learning methods, especially deep learning methods, rely heavily on a large number of annotation data, which human cognition often does not need. This weakness greatly limits the application of deep learning method in practical problem. To solve this problem, learning from few shot examples attracts more and more community''s research interest. In order to better understand the few shot learning problem, this paper extensively discusses several popular few shot learning methods, including data augmentation methods, transfer learning methods and meta learning methods. By discussing the processes and core ingredients of different algorithms, we can clearly see the advantages and disadvantages of existing methods in solving few shot learning problems. At the end of this paper, we highlight the points to future research directions in the field of few shot learning problem.