Deep Reinforcement Learning with Label Embedding Reward for Supervised Image Hashing
This work addresses image retrieval efficiency for applications like large-scale search, presenting a novel approach rather than an incremental improvement.
The paper tackles the problem of generating binary codes for image retrieval by formulating hashing as a path-finding task in binary space and using deep reinforcement learning with a label embedding reward based on BCH codes. It reports outperforming state-of-the-art supervised hashing methods on CIFAR-10 and NUS-WIDE datasets across various code lengths.
Deep hashing has shown promising results in image retrieval and recognition. Despite its success, most existing deep hashing approaches are rather similar: either multi-layer perceptron or CNN is applied to extract image feature, followed by different binarization activation functions such as sigmoid, tanh or autoencoder to generate binary code. In this work, we introduce a novel decision-making approach for deep supervised hashing. We formulate the hashing problem as travelling across the vertices in the binary code space, and learn a deep Q-network with a novel label embedding reward defined by Bose-Chaudhuri-Hocquenghem (BCH) codes to explore the best path. Extensive experiments and analysis on the CIFAR-10 and NUS-WIDE dataset show that our approach outperforms state-of-the-art supervised hashing methods under various code lengths.