In Defense of LSTMs for Addressing Multiple Instance Learning Problems

TL;DR

This paper demonstrates that LSTMs are highly effective for multiple instance learning tasks, even with unordered data, and can learn instance-level information using only bag-level labels, outperforming some specialized methods.

Contribution

It shows that LSTMs can be effectively applied to MIL problems, capturing instance-level info with weak supervision, which is a novel application for this model type.

Findings

LSTMs perform well on MIL tasks across various datasets.

LSTMs can learn instance-level predictions with only bag-level annotations.

Their performance approaches that of fully-supervised methods.

Abstract

LSTMs have a proven track record in analyzing sequential data. But what about unordered instance bags, as found under a Multiple Instance Learning (MIL) setting? While not often used for this, we show LSTMs excell under this setting too. In addition, we show thatLSTMs are capable of indirectly capturing instance-level information us-ing only bag-level annotations. Thus, they can be used to learn instance-level models in a weakly supervised manner. Our empirical evaluation on both simplified (MNIST) and realistic (Lookbook and Histopathology) datasets shows that LSTMs are competitive with or even surpass state-of-the-art methods specially designed for handling specific MIL problems. Moreover, we show that their performance on instance-level prediction is close to that of fully-supervised methods.