We introduce an online action recognition system that can be combined with any set of frame-by-frame feature descriptors. Our system covers the frame feature space with classifiers whose distribution adapts to the hardness of locally approximating the Bayes optimal classifier. An efficient nearest neighbour search is used to find and combine the local classifiers that are closest to the frames of a new video to be classified. The advantages of our approach are: incremental training, frame by frame real-time prediction, nonparametric predictive modelling, video segmentation for continuous action recognition, no need to trim videos to equal lengths and only one tuning parameter (which, for large datasets, can be safely set to the diameter of the feature space). Experiments on standard benchmarks show that our system is competitive with state-of-the-art non incremental and incremental baselines. keywords: action recognition, incremental learning, continuous action recognition, nonparametric model, real time, multivariate time series classification, temporal classification.
Online action recognition via nonparametric incremental learning / R. De Rosa, N. Cesa-Bianchi, I. Gori, F. Cuzzolin - In: Proceedings of the British Machine Vision Conference / [a cura di] M. Valstar, A. French, T. Pridmore. - [s.l] : BMVA Press, 2014 Sep. - ISBN 1901725529. - pp. 1-15 (( Intervento presentato al 25. convegno British Machine Vision Conference tenutosi a Nottingham nel 2014 [10.5244/C.28.113].
Online action recognition via nonparametric incremental learning
R. De Rosa;N. Cesa-Bianchi;
2014
Abstract
We introduce an online action recognition system that can be combined with any set of frame-by-frame feature descriptors. Our system covers the frame feature space with classifiers whose distribution adapts to the hardness of locally approximating the Bayes optimal classifier. An efficient nearest neighbour search is used to find and combine the local classifiers that are closest to the frames of a new video to be classified. The advantages of our approach are: incremental training, frame by frame real-time prediction, nonparametric predictive modelling, video segmentation for continuous action recognition, no need to trim videos to equal lengths and only one tuning parameter (which, for large datasets, can be safely set to the diameter of the feature space). Experiments on standard benchmarks show that our system is competitive with state-of-the-art non incremental and incremental baselines. keywords: action recognition, incremental learning, continuous action recognition, nonparametric model, real time, multivariate time series classification, temporal classification.File | Dimensione | Formato | |
---|---|---|---|
paper110.pdf
accesso riservato
Tipologia:
Publisher's version/PDF
Dimensione
1.38 MB
Formato
Adobe PDF
|
1.38 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.