The motivation is to train deep cnn for image retrieval. Potential candidate losses are weighted approximate ranking (WARP)[1] and SVM[2]. [1] [Yunchao Gong, Yangqing Jia, Sergey Ioffe, Alexander Toshev, Thomas Leung. Deep Convolutional Ranking for Multilabel Image Annotation. arXiv:1312.4894 [cs.CV]](http://openreview.net/document/52215b11-6716-4899-bce5-4479b49dc379#52215b11-6716-4899-bce5-4479b49dc379) [2] [Wei Yu, Tiejun Zhao, Yalong Bai, Wei-Ying Ma, Kuiyuan Yang. Learning High-level Image Representation for Image Retrieval via Multi-Task DNN using Clickthrough Data. arXiv:1312.4740 [cs.CV]](http://openreview.net/document/90fc8dad-ad02-4ddc-ab06-e7b55706869d#90fc8dad-ad02-4ddc-ab06-e7b55706869d)
The motivation is to train deep cnn for image retrieval. Potential candidate losses are weighted approximate ranking (WARP)[1] and SVM[2].
[1] Yunchao Gong, Yangqing Jia, Sergey Ioffe, Alexander Toshev, Thomas Leung. Deep Convolutional Ranking for Multilabel Image Annotation. arXiv:1312.4894 [cs.CV]
[2] Wei Yu, Tiejun Zhao, Yalong Bai, Wei-Ying Ma, Kuiyuan Yang. Learning High-level Image Representation for Image Retrieval via Multi-Task DNN using Clickthrough Data. arXiv:1312.4740 [cs.CV]