DVHN: A Deep Hashing Framework for Large-Scale Vehicle Re-Identification

Abstract

Vehicle re-identification is a pervasive technology in real-world intelligence transportation systems. Conventional methods generally perform re-identification tasks by representing vehicle images as real-valued feature vectors and then ranking the gallery images by computing the corresponding Euclidean distances. Despite achieving remarkable retrieval accuracy, these high-dimensional real-valued feature vectors are not tailored for fast indexing and matching and require tremendous memory and computation when the gallery set is large, making them inapplicable in a large-scale real-world retrieval setting. In light of this limitation, in this paper, we make the very first attempt to develop an efficient vehicle re-identification system ( <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DVHN</b> ) for real-world large-scale retrieval tasks with deep hashing learning. It could substantially reduce memory usage and enhances retrieval efficiency while maintaining retrieval accuracy. Concretely, DVHN directly learns discrete compact binary hashing codes for each image by jointly optimizing the feature learning network and the hash code generating module. Specifically, we directly constrain the output from the convolutional neural network to be discrete binary codes and ensure the learned binary codes are optimal for classification. To optimize the deep discrete hashing framework, we further propose an alternating minimization method for learning binary similarity-preserved hashing codes. Extensive experiments on two widely-studied vehicle re-identification datasets- <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">VehicleID</b> and <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">VeRi</b> -have demonstrated the superiority of our method against the state-of-the-art deep hash methods. <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DVHN</b> of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$2048$</tex-math> </inline-formula> bits can achieve 13.94% and 10.21% accuracy improvement in terms of <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">mAP</b> and <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Rank@1</b> for <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">VehicleID (800)</b> dataset. For <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">VeRi</b> , we achieve 35.45% and 32.72% performance gains for <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Rank@1</b> and <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">mAP</b> , respectively.