Date of Original Version

9-2014

Type

Conference Proceeding

Journal Title

Lecture Notes in Computer Science

Volume

8689

First Page

593

Last Page

608

Rights Management

The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-319-10593-2_39

Abstract or Description

An increasing number of computer vision and pattern recognition problems require structured regression techniques. Problems like human pose estimation, unsegmented action recognition, emotion prediction and facial landmark detection have temporal or spatial output dependencies that regular regression techniques do not capture. In this paper we present continuous conditional neural fields (CCNF) – a novel structured regression model that can learn non-linear input-output dependencies, and model temporal and spatial output relationships of varying length sequences. We propose two instances of our CCNF framework: Chain-CCNF for time series modelling, and Grid-CCNF for spatial relationship modelling. We evaluate our model on five public datasets spanning three different regression problems: facial landmark detection in the wild, emotion prediction in music and facial action unit recognition. Our CCNF model demonstrates state-of-the-art performance on all of the datasets used.

DOI

10.1007/978-3-319-10593-2_39

Share

COinS
 

Published In

Lecture Notes in Computer Science, 8689, 593-608.