| With the emergence of machine learning,the era of artificial intelligence is coming.In recent years,the development of deep learning technology has attracted world-wide attention,especially in computer vision technology.Image similarity learning has achieved good development in various fields of society.Learning fine-grained image similarity is a challenging task in machine learning,especially for highly similar images.In this paper,a novel image similarity learning method based on triplet loss function and capsule network is proposed.The triplet loss can reflect the "relative" similarity between images,while capsule network has better image feature extraction ability than common deep learning network.Combining them together can improve image similarity learning results.The experimental results show that the accuracy of the new method proposed in this paper is much higher than that of the previous deep learning method. |
