On Trivial Solution and High Correlation Problems in Deep Supervised Hashing

Guo, Yuchen and Zhao, Xin and Ding, Guiguang and Han, Jungong (2018) On Trivial Solution and High Correlation Problems in Deep Supervised Hashing. In: 32nd AAAI Conference on Artificial Intelligence 2018. AAAI, Palo Alto, pp. 2240-2247. ISBN 9781577358008

[img]
Preview
PDF (2018-3)
2018_3.pdf - Accepted Version

Download (2MB)

Abstract

Deep supervised hashing (DSH), which combines binary learning and convolutional neural network, has attracted considerable research interests and achieved promising performance for highly efficient image retrieval. In this paper, we show that the widely used loss functions, pair-wise loss and triplet loss, suffer from the trivial solution problem and usually lead to highly correlated bits in practice, limiting the performance of DSH. One important reason is that it is difficult to incorporate proper constraints into the loss functions under the mini-batch based optimization algorithm. To tackle these problems, we propose to adopt ensemble learning strategy for deep model training. We found out that this simple strategy is capable of effectively decorrelating different bits, making the hashcodes more informative. Moreover, it is very easy to parallelize the training and support incremental model learning, which are very useful for real-world applications but usually ignored by existing DSH approaches. Experiments on benchmarks demonstrate the proposed ensemble based DSH can improve the performance of DSH approaches significant.

Item Type: Contribution in Book/Report/Proceedings
Departments: Faculty of Science and Technology > School of Computing & Communications
ID Code: 123575
Deposited By: ep_importer_pure
Deposited On: 22 Feb 2018 16:40
Refereed?: Yes
Published?: Published
Last Modified: 25 Feb 2020 05:22
URI: https://eprints.lancs.ac.uk/id/eprint/123575

Actions (login required)

View Item View Item