Introspective Deep Metric Learning

TL;DR

This paper introduces an introspective deep metric learning framework that incorporates uncertainty modeling into image similarity assessments, leading to more robust and accurate image retrieval and clustering.

Contribution

It proposes a novel IDML framework that represents images with both semantic and uncertainty embeddings, and an introspective similarity metric that considers uncertainties during learning.

Findings

Achieves state-of-the-art results on multiple datasets.

Improves robustness by modeling uncertainty in image representations.

Enables adaptive learning pace considering uncertainty.

Abstract

This paper proposes an introspective deep metric learning (IDML) framework for uncertainty-aware comparisons of images. Conventional deep metric learning methods focus on learning a discriminative embedding to describe the semantic features of images, which ignore the existence of uncertainty in each image resulting from noise or semantic ambiguity. Training without awareness of these uncertainties causes the model to overfit the annotated labels during training and produce unsatisfactory judgments during inference. Motivated by this, we argue that a good similarity model should consider the semantic discrepancies with awareness of the uncertainty 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 gradient analysis of the proposed metric shows that it enables the model to learn at an adaptive and slower pace to deal with the uncertainty during training. 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 Products datasets for image retrieval and clustering. We further provide an in-depth analysis of our framework to demonstrate the effectiveness and reliability of IDML. Code: https://github.com/wzzheng/IDML.