Local Aggregation for Unsupervised Learning of Visual Embeddings

TL;DR

This paper introduces a local aggregation method for training neural network embeddings in an unsupervised manner, improving large-scale visual recognition tasks by enabling the emergence of soft clusters and achieving state-of-the-art transfer learning results.

Contribution

It presents a novel local aggregation training approach that enhances unsupervised visual embedding learning by dynamically forming soft clusters, leading to improved recognition performance.

Findings

Achieved state-of-the-art unsupervised transfer learning on ImageNet.

Improved scene recognition on Places 205.

Enhanced object detection on PASCAL VOC.

Abstract

Unsupervised approaches to learning in neural networks are of substantial interest for furthering artificial intelligence, both because they would enable the training of networks without the need for large numbers of expensive annotations, and because they would be better models of the kind of general-purpose learning deployed by humans. However, unsupervised networks have long lagged behind the performance of their supervised counterparts, especially in the domain of large-scale visual recognition. Recent developments in training deep convolutional embeddings to maximize non-parametric instance separation and clustering objectives have shown promise in closing this gap. Here, we describe a method that trains an embedding function to maximize a metric of local aggregation, causing similar data instances to move together in the embedding space, while allowing dissimilar instances to separate. This aggregation metric is dynamic, allowing soft clusters of different scales to emerge. We evaluate our procedure on several large-scale visual recognition datasets, achieving state-of-the-art unsupervised transfer learning performance on object recognition in ImageNet, scene recognition in Places 205, and object detection in PASCAL VOC.