A weakly supervised adaptive triplet loss for deep metric learning

TL;DR

This paper introduces a weakly supervised adaptive triplet loss for deep metric learning that leverages weak labels to improve image embedding quality, especially across different domains, without extensive annotation.

Contribution

It proposes a novel weakly supervised adaptive triplet loss that captures fine-grained semantic similarity and enhances cross-domain generalization in image retrieval tasks.

Findings

Boosts triplet loss performance by 10.6% on cross-domain data

Outperforms state-of-the-art models on benchmark datasets

Effectively uses weakly labeled data to learn semantic classes

Abstract

We address the problem of distance metric learning in visual similarity search, defined as learning an image embedding model which projects images into Euclidean space where semantically and visually similar images are closer and dissimilar images are further from one another. We present a weakly supervised adaptive triplet loss (ATL) capable of capturing fine-grained semantic similarity that encourages the learned image embedding models to generalize well on cross-domain data. The method uses weakly labeled product description data to implicitly determine fine grained semantic classes, avoiding the need to annotate large amounts of training data. We evaluate on the Amazon fashion retrieval benchmark and DeepFashion in-shop retrieval data. The method boosts the performance of triplet loss baseline by 10.6% on cross-domain data and out-performs the state-of-art model on all evaluation metrics.