Deep unsupervised learning through spatial contrasting

Abstract

Convolutional networks have marked their place over the last few years as the best performing model for various visual tasks. They are, however, most suited for supervised learning from large amounts of labeled data. Previous attempts have been made to use unlabeled data to improve model performance by applying unsupervised techniques. These attempts require different architectures and training methods. In this work we present a novel approach for unsupervised training of Convolutional networks that is based on contrasting between spatial regions within images. This criterion can be employed within conventional neural networks and trained using standard techniques such as SGD and back-propagation, thus complementing supervised methods.

Cited by in corpus (8)

• Matterport3D: Learning from RGB-D Data in Indoor Environments
• Regularization for Deep Learning: A Taxonomy
• Self-supervised Visual Feature Learning with Deep Neural Networks: A Survey
• Self-Supervised Spatiotemporal Feature Learning via Video Rotation Prediction
• Scattering Networks for Hybrid Representation Learning
• Robotic Pick-and-Place of Novel Objects in Clutter with Multi-Affordance Grasping and Cross-Domain Image Matching
• Multimodal Classification for Analysing Social Media
• Scaling the Scattering Transform: Deep Hybrid Networks