Structured Prediction Energy Networks

Abstract

We introduce structured prediction energy networks (SPENs), a flexible framework for structured prediction. A deep architecture is used to define an energy function of candidate labels, and then predictions are produced by using back-propagation to iteratively optimize the energy with respect to the labels. This deep architecture captures dependencies between labels that would lead to intractable graphical models, and performs structure learning by automatically learning discriminative features of the structured output. One natural application of our technique is multi-label classification, which traditionally has required strict prior assumptions about the interactions between labels to ensure tractable learning and prediction. We are able to apply SPENs to multi-label problems with substantially larger label sets than previous applications of structured prediction, while modeling high-order interactions using minimal structural assumptions. Overall, deep learning provides remarkable tools for learning features of the inputs to a prediction problem, and this work extends these techniques to learning features of structured outputs. Our experiments provide impressive performance on a variety of benchmark multi-label classification tasks, demonstrate that our technique can be used to provide interpretable structure learning, and illuminate fundamental trade-offs between feed-forward and iterative structured prediction.

ICML 2016

References in corpus (9)

• Sequence to Sequence Learning with Neural Networks
• Natural Language Processing (almost) from Scratch
• Grammar as a Foreign Language
• Gradient-based Hyperparameter Optimization through Reversible Learning
• Multi-Label Prediction via Compressed Sensing
• Fully Connected Deep Structured Networks
• Deep Unfolding: Model-Based Inspiration of Novel Deep Architectures
• Deep Structured Output Learning for Unconstrained Text Recognition
• Locally Non-linear Embeddings for Extreme Multi-label Learning

Cited by in corpus (9)

• B-CNN: Branch Convolutional Neural Network for Hierarchical Classification
• Personalized Bundle List Recommendation
• Detecting Visual Relationships with Deep Relational Networks
• Learning deep structured active contours end-to-end
• CERN: Confidence-Energy Recurrent Network for Group Activity Recognition
• The Limited Multi-Label Projection Layer
• Learning to Search on Manifolds for 3D Pose Estimation of Articulated Objects
• End-to-end learning potentials for structured attribute prediction
• Learning Output Embeddings in Structured Prediction