One Loss For Quantization: Deep Hashing With Discrete Wasserstein Distributional Matching

Abstract

Image hashing is a principled approximate nearest neighbor approach to find similar items to a query in a large collection of images. Hashing aims to learn a binary-output function that maps an image to a binary vector. For optimal retrieval performance, producing balanced hash codes with low-quantization error to bridge the gap between the learning stage's continuous relaxation and the inference stage's discrete quantization is important. However, in the existing deep supervised hashing methods, coding balance and low-quantization error are difficult to achieve and involve several losses. We argue that this is because the existing quantization approaches in these methods are heuristically constructed and not effective to achieve these objectives. This paper considers an alternative approach to learning the quantization constraints. The task of learning balanced codes with low quantization error is re-formulated as matching the learned distribution of the continuous codes to a pre-defi

Stats

• citations44
• S2 citations—
• github stars0
• HF likes0
• heat score12.40
• arxiv keydoan2022one

Related papers

• Deep Hashing Via Householder Quantization (2023)0.00
• Push For Quantization: Deep Fisher Hashing (2019)0.00
• Simultaneous Compression And Quantization: A Joint Approach For Efficient Unsupervised Hashing (2018)7.50
• Deep Priority Hashing (2018)11.67
• Deep Hash Distillation For Image Retrieval (2021)11.29
• HHF: Hashing-guided Hinge Function For Deep Hashing Retrieval (2021)14.24
• One Loss For All: Deep Hashing With A Single Cosine Similarity Based Learning Objective (2021)3.99
• Deep Supervised Discrete Hashing (2017)0.00