Weakly Supervised Representation Learning with Coarse Labels

TL;DR

This paper introduces a novel algorithm for learning fine-grained representations from coarse labels, backed by theoretical guarantees and validated through extensive experiments, addressing challenges in real-world applications with limited task-specific labels.

Contribution

The paper proposes a new method to learn detailed representations from coarse labels, providing theoretical guarantees and demonstrating significant improvements in real-world datasets.

Findings

Improved representation quality on target tasks with coarse labels

Theoretical guarantee for the proposed learning algorithm

Significant performance gains demonstrated through experiments

Abstract

With the development of computational power and techniques for data collection, deep learning demonstrates a superior performance over most existing algorithms on visual benchmark data sets. Many efforts have been devoted to studying the mechanism of deep learning. One important observation is that deep learning can learn the discriminative patterns from raw materials directly in a task-dependent manner. Therefore, the representations obtained by deep learning outperform hand-crafted features significantly. However, for some real-world applications, it is too expensive to collect the task-specific labels, such as visual search in online shopping. Compared to the limited availability of these task-specific labels, their coarse-class labels are much more affordable, but representations learned from them can be suboptimal for the target task. To mitigate this challenge, we propose an algorithm to learn the fine-grained patterns for the target task, when only its coarse-class labels are available. More importantly, we provide a theoretical guarantee for this. Extensive experiments on real-world data sets demonstrate that the proposed method can significantly improve the performance of learned representations on the target task, when only coarse-class information is available for training. Code is available at \url{https://github.com/idstcv/CoIns}.