Learning A Deep l∞ Encoder for Hashing

Wang Zhangyang, Yang Yingzhen, Chang Shiyu, Ling Qing, Thomas S. Huang

Research output: Contribution to journalConference articlepeer-review

11 Scopus citations


We investigate the l∞-constrained representation which demonstrates robustness to quantization errors, utilizing the tool of deep learning. Based on the Alternating Direction Method of Multipliers (ADMM), we formulate the original convex minimization problem as a feed-forward neural network, named Deep l∞ Encoder, by introducing the novel Bounded Linear Unit (BLU) neuron and modeling the Lagrange multipliers as network biases. Such a structural prior acts as an effective network regularization, and facilitates the model initialization. We then investigate the effective use of the proposed model in the application of hashing, by coupling the proposed encoders under a supervised pairwise loss, to develop a Deep Siamese Network, which can be optimized from end to en1d. Extensive experiments demonstrate the impressive performances of the proposed model. We also provide an in-depth analysis of its behaviors against the competitors.

Original languageEnglish (US)
Pages (from-to)2174-2180
Number of pages7
JournalIJCAI International Joint Conference on Artificial Intelligence
StatePublished - 2016
Externally publishedYes
Event25th International Joint Conference on Artificial Intelligence, IJCAI 2016 - New York, United States
Duration: Jul 9 2016Jul 15 2016

ASJC Scopus subject areas

  • Artificial Intelligence


Dive into the research topics of 'Learning A Deep l∞ Encoder for Hashing'. Together they form a unique fingerprint.

Cite this