Open AccessSemi‐supervised long short‐term memory for human action recognition
Author(s) -
Liu Hong,
Liu Chang,
Ding Runwei
Publication year - 2020
Publication title -
the journal of engineering
Language(s) - English
Resource type - Journals
ISSN - 2051-3305
DOI - 10.1049/joe.2019.1166
Subject(s) - discriminative model , computer science , artificial intelligence , term (time) , task (project management) , action recognition , long short term memory , machine learning , action (physics) , pattern recognition (psychology) , focus (optics) , rgb color model , artificial neural network , recurrent neural network , engineering , class (philosophy) , physics , quantum mechanics , systems engineering , optics
In real human action recognition task, it is a common phenomenon that there are many unlabelled samples and few labelled samples. How to make good use of unlabelled samples to improve the generalisation ability of models is the focus of semi‐supervised learning research. In this study, the authors present two semi‐supervised methods based on long short‐term memory (LSTM) to learn discriminative hidden features. One is the LSTM ladder network, the other is the Symmetrical LSTM network. By them unlabelled samples can be used automatically to improve learning performance without relying on external interaction. Both on the NTU‐RGB + D dataset and the Kinetics dataset, their methods achieve >10 and 5% improvements, separately.