@inproceedings{b92b10e6c5fe40ed948ea9f42e6ef145,
title = "User identification and object recognition in clutter scenes based on RGB-depth analysis",
abstract = "We propose an automatic system for user identification and object recognition based on multi-modal RGB-Depth data analysis. We model a RGBD environment learning a pixel-based background Gaussian distribution. Then, user and object candidate regions are detected and recognized online using robust statistical approaches over RGBD descriptions. Finally, the system saves the historic of user-object assignments, being specially useful for surveillance scenarios. The system has been evaluated on a novel data set containing different indoor/outdoor scenarios, objects, and users, showing accurate recognition and better performance than standard state-of-the-art approaches.",
keywords = "Multi-modal RGB-Depth data analysis, Object Recognition, Statistical learning, User identification, Visual features",
author = "Albert Clap{\'e}s and Miguel Reyes and Sergio Escalera",
year = "2012",
doi = "10.1007/978-3-642-31567-1_1",
language = "English",
isbn = "9783642315664",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Physica-Verlag",
pages = "1--11",
booktitle = "Articulated Motion and Deformable Objects - 7th International Conference, AMDO 2012, Proceedings",
note = "7th International Conference on Articulated Motion and Deformable Objects, AMDO 2012 ; Conference date: 11-07-2012 Through 13-07-2012",
}