Please use this identifier to cite or link to this item:
|Title:||Deeply learning the messages in message passing inference|
Van Den Hengel, A.
|Citation:||Advances in Neural Information Processing Systems 28: 29th Annual Conference on Neural Information Processing Systems 2015, 2015 / Cortes, C., Lawrence, N., Lee, D., Sugiyama, M., Garnett, R. (ed./s), vol.2015-January, pp.361-369|
|Publisher:||Neural Information Processing Systems|
|Series/Report no.:||Advances in Neural Information Processing Systems|
|Conference Name:||29th Annual Conference on Neural Information Processing Systems 2015 (NIPS 2015) (07 Dec 2015 - 12 Dec 2015 : Montreal)|
|Guosheng Lin, Chunhua Shen, Ian Reid, Anton van den Hengel|
|Abstract:||Deep structured output learning shows great promise in tasks like semantic image segmentation. We proffer a new, efficient deep structured model learning scheme, in which we show how deep Convolutional Neural Networks (CNNs) can be used to directly estimate the messages in message passing inference for structured prediction with Conditional Random Fields CRFs). With such CNN message estimators, we obviate the need to learn or evaluate potential functions for message calculation. This confers significant efficiency for learning, since otherwise when performing structured learning for a CRF with CNN potentials it is necessary to undertake expensive inference for every stochastic gradient iteration. The network output dimension of message estimators is the same as the number of classes, rather than exponentially growing in the order of the potentials. Hence it is more scalable for cases that a large number of classes are involved. We apply our method to semantic image segmentation and achieve impressive performance, which demonstrates the effectiveness and usefulness of our CNN message learning method.|
|Rights:||Copyright © (2015) by Neural Information Processing Systems All rights reserved.|
|Appears in Collections:||Australian Institute for Machine Learning publications|
Computer Science publications
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.