Skip to yearly menu bar Skip to main content


Poster

Learning Deep Parsimonious Representations

Renjie Liao · Alex Schwing · Richard Zemel · Raquel Urtasun

Area 5+6+7+8 #169

Keywords: [ Deep Learning or Neural Networks ] [ Clustering ]


Abstract:

In this paper we aim at facilitating generalization for deep networks while supporting interpretability of the learned representations. Towards this goal, we propose a clustering based regularization that encourages parsimonious representations. Our k-means style objective is easy to optimize and flexible supporting various forms of clustering, including sample and spatial clustering as well as co-clustering. We demonstrate the effectiveness of our approach on the tasks of unsupervised learning, classification, fine grained categorization and zero-shot learning.

Live content is unavailable. Log in and register to view live content