Introspective Deep Metric Learning For Image Retrieval

Abstract

This paper proposes an introspective deep metric learning (IDML) framework for uncertainty-aware comparisons of images. Conventional deep metric learning methods produce confident semantic distances between images regardless of the uncertainty level. However, we argue that a good similarity model should consider the semantic discrepancies with caution to better deal with ambiguous images for more robust training. To achieve this, we propose to represent an image using not only a semantic embedding but also an accompanying uncertainty embedding, which describes the semantic characteristics and ambiguity of an image, respectively. We further propose an introspective similarity metric to make similarity judgments between images considering both their semantic differences and ambiguities. The proposed IDML framework improves the performance of deep metric learning through uncertainty modeling and attains state-of-the-art results on the widely used CUB-200-2011, Cars196, and Stanford Online

Stats

• citations0
• S2 citations—
• github stars55
• HF likes0
• heat score3.50
• arxiv keyzheng2022introspective

Related papers

• Introspective Deep Metric Learning (2023)13.85
• Hyp-uml: Hyperbolic Image Retrieval With Uncertainty-aware Metric Learning (2023)0.00
• Deep Relational Metric Learning (2021)12.17
• Three Things To Know About Deep Metric Learning (2024)0.00
• Unbiased Evaluation Of Deep Metric Learning Algorithms (2019)0.00
• Indirect: Language-guided Zero-shot Deep Metric Learning For Images (2022)5.24
• Directional Statistics-based Deep Metric Learning For Image Classification And Retrieval (2018)13.05
• Large-scale Distance Metric Learning With Uncertainty (2018)9.03