In this paper we present an efficient method for visual descriptors retrievalbased on compact hash codes computed using a multiple k-means assignment. Themethod has been applied to the problem of approximate nearest neighbor (ANN)search of local and global visual content descriptors, and it has been testedon different datasets: three large scale public datasets of up to one billiondescriptors (BIGANN) and, supported by recent progress in convolutional neuralnetworks (CNNs), also on the CIFAR-10 and MNIST datasets. Experimental resultsshow that, despite its simplicity, the proposed method obtains a very highperformance that makes it superior to more complex state-of-the-art methods.
展开▼
