Binary Generative Adversarial Networks for Image Retrieval

Abstract

The most striking successes in image retrieval using deep hashing have mostly involved discriminative models, which require labels. In this paper, we use binary generative adversarial networks (BGAN) to embed images to binary codes in an unsupervised way. By restricting the input noise variable of generative adversarial networks (GAN) to be binary and conditioned on the features of each input image, BGAN can simultaneously learn a binary representation per image, and generate an image plausibly similar to the original one. In the proposed framework, we address two main problems: 1) how to directly generate binary codes without relaxation? 2) how to equip the binary representation with the ability of accurate image retrieval? We resolve these problems by proposing new sign-activation strategy and a loss function steering the learning process, which consists of new models for adversarial loss, a content loss, and a neighborhood structure loss. Experimental results on standard datasets (CIFAR-10, NUSWIDE, and Flickr) demonstrate that our BGAN significantly outperforms existing hashing methods by up to 107\% in terms of~mAP (See Table tab.res.map.comp) Our anonymous code is available at: https://github.com/htconquer/BGAN.

arXiv admin note: text overlap with arXiv:1702.00758 by other authors

Cited by in corpus (17)

• A Decade Survey of Content Based Image Retrieval using Deep Learning
• A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications
• DML-GANR: Deep Metric Learning With Generative Adversarial Network Regularization for High Spatial Resolution Remote Sensing Image Retrieval
• From Selective Deep Convolutional Features to Compact Binary Representations for Image Retrieval
• Embarrassingly Simple Binary Representation Learning
• Auto-Encoding Twin-Bottleneck Hashing
• Two-pronged Strategy: Lightweight Augmented Graph Network Hashing for Scalable Image Retrieval
• New Ideas and Trends in Deep Multimodal Content Understanding: A Review
• Simultaneous Compression and Quantization: A Joint Approach for Efficient Unsupervised Hashing
• Learning Product Codebooks using Vector Quantized Autoencoders for Image Retrieval
• Excavate Condition-invariant Space by Intrinsic Encoder
• Enhance Gender and Identity Preservation in Face Aging Simulation for Infants and Toddlers
• Online Hashing with Efficient Updating of Binary Codes
• Unsupervised Domain-adaptive Hash for Networks
• A Domain-Oblivious Approach for Learning Concise Representations of Filtered Topological Spaces for Clustering
• Deep Self-Adaptive Hashing for Image Retrieval
• Fast Search on Binary Codes by Weighted Hamming Distance