The detection of human-object interaction features is a valuable tool in fields such as developmental psychology. However, the traditional method of manual labeling by psychologists is time-consuming and labor-intensive. Current interaction recognition methods in computer vision are limited to single-angle images, which are susceptible to occlusions and not suitable for continuous prediction. In this study, we propose a novel video-based interactor detection method for interaction partners and manipulated objects that employs a multiview processing stage that extracts the optimal angle to account for occlusions and ensure the continuity of predictions. We validated the proposed method using videos of four pairs of interactors obtained from three angles. The experimental results demonstrate the effectiveness of the proposed method, which achieved an average accuracy of 82.3% on the video test set.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.