# Learning from Web Data with Self-Organizing Memory Module

## TL;DR

This paper introduces a novel end-to-end trainable method that effectively handles label and background noise in web images using a self-organizing memory module within a multi-instance learning framework, improving performance on benchmark datasets.

## Contribution

The proposed approach uniquely integrates a self-organizing memory module with multi-instance learning to manage noisy web data without requiring clean image supervision.

## Key findings

- Outperforms existing methods on four benchmark datasets
- Effectively handles label and background noise simultaneously
- Integrates memory module seamlessly with classification for end-to-end training

## Abstract

Learning from web data has attracted lots of research interest in recent years. However, crawled web images usually have two types of noises, label noise and background noise, which induce extra difficulties in utilizing them effectively. Most existing methods either rely on human supervision or ignore the background noise. In this paper, we propose a novel method, which is capable of handling these two types of noises together, without the supervision of clean images in the training stage. Particularly, we formulate our method under the framework of multi-instance learning by grouping ROIs (i.e., images and their region proposals) from the same category into bags. ROIs in each bag are assigned with different weights based on the representative/discriminative scores of their nearest clusters, in which the clusters and their scores are obtained via our designed memory module. Our memory module could be naturally integrated with the classification module, leading to an end-to-end trainable system. Extensive experiments on four benchmark datasets demonstrate the effectiveness of our method.