Action classification in soccer videos with long short-term memory recurrent neural networks

Authors:
Moez Baccouche;Franck Mamalet;Christian Wolf;Christophe Garcia;Atilla Baskurt
Affiliations:
Orange Labs, Cesson-Sévigné, France and LIRIS, UMR, CNRS, INSA-Lyon, France;Orange Labs, Cesson-Sévigné, France;LIRIS, UMR, CNRS, INSA-Lyon, France;Orange Labs, Cesson-Sévigné, France;LIRIS, UMR, CNRS, INSA-Lyon, France
Venue:
ICANN'10 Proceedings of the 20th international conference on Artificial neural networks: Part II
Year:
2010

Citing 6
Cited 1

Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography

Communications of the ACM
Automatic Parsing of TV Soccer Programs

ICMCS '95 Proceedings of the International Conference on Multimedia Computing and Systems
Learning precise timing with lstm recurrent networks

The Journal of Machine Learning Research
Distinctive Image Features from Scale-Invariant Keypoints

International Journal of Computer Vision
Action Categorization in Soccer Videos Using String Kernels

CBMI '09 Proceedings of the 2009 Seventh International Workshop on Content-Based Multimedia Indexing
Automatic soccer video analysis and summarization

IEEE Transactions on Image Processing

Sequential deep learning for human action recognition

HBU'11 Proceedings of the Second international conference on Human Behavior Unterstanding

Quantified Score

Hi-index	0.01

Visualization

Abstract

In this paper, we propose a novel approach for action classification in soccer videos using a recurrent neural network scheme. Thereby, we extract from each video action at each timestep a set of features which describe both the visual content (by the mean of a BoW approach) and the dominant motion (with a key point based approach). A Long Short-Term Memory-based Recurrent Neural Network is then trained to classify each video sequence considering the temporal evolution of the features for each timestep. Experimental results on the MICC-Soccer-Actions-4 database show that the proposed approach outperforms classification methods of related works (with a classification rate of 77%), and that the combination of the two features (BoW and dominant motion) leads to a classification rate of 92%.