A Lower Bound Of Hash Codes' Performance

Abstract

As a crucial approach for compact representation learning, hashing has achieved great success in effectiveness and efficiency. Numerous heuristic Hamming space metric learning objectives are designed to obtain high-quality hash codes. Nevertheless, a theoretical analysis of criteria for learning good hash codes remains largely unexploited. In this paper, we prove that inter-class distinctiveness and intra-class compactness among hash codes determine the lower bound of hash codes' performance. Promoting these two characteristics could lift the bound and improve hash learning. We then propose a surrogate model to fully exploit the above objective by estimating the posterior of hash codes and controlling it, which results in a low-bias optimization. Extensive experiments reveal the effectiveness of the proposed method. By testing on a series of hash-models, we obtain performance improvements among all of them, with an up to \(26.5%\) increase in mean Average Precision and an up to \(20.5%

Stats

• citations0
• S2 citations—
• github stars5
• HF likes0
• heat score1.56
• arxiv keyzhu2022a

Related papers

• Collaborative Learning For Extremely Low Bit Asymmetric Hashing (2018)7.50
• Compact Hash Code Learning With Binary Deep Neural Network (2017)9.03
• How Should We Evaluate Supervised Hashing? (2016)12.68
• CIMON: Towards High-quality Hash Codes (2020)10.61
• Learning Hash Function Through Codewords (2019)0.00
• Simultaneous Feature Aggregating And Hashing For Compact Binary Code Learning (2019)9.92
• Embarrassingly Simple Binary Representation Learning (2019)9.41
• Learning To Hash With Semantic Similarity Metrics And Empirical KL Divergence (2020)0.00