A New Hybrid Architecture for Human Activity Recognition from RGB-D videos

Srijan Das (1) , Monique Thonnat (1) , Kaustubh Sakhalkar (1) , Michal F Koperski (1) , Francois Bremond (1) , Gianpiero Francesca (2)

1 STARS - Spatio-Temporal Activity Recognition Systems
2 Toyota Motor Europe

Activity Recognition from RGB-D videos is still an open problem due to the presence of large varieties of actions. In this work, we propose a new architecture by mixing a high level handcrafted strategy and machine learning techniques. We propose a novel two level fusion strategy to combine features from different cues to address the problem of large variety of actions. As similar actions are common in daily living activities, we also propose a mechanism for similar action discrimination. We validate our approach on four public datasets, CAD-60, CAD-120, MSRDailyActivity3D, and NTU-RGB+D improving the state-of-the-art results on them.

Mots clés: Activity recognition; RGB-D videos; Data fusion

Domaines: Informatique [cs]

Srijan Das, Monique Thonnat, Kaustubh Sakhalkar, Michal F Koperski, Francois Bremond, et al.. A New Hybrid Architecture for Human Activity Recognition from RGB-D videos. MMM 2019 - 25th International Conference on MultiMedia Modeling, Jan 2019, Thessaloniki, Greece. pp.493-505, ⟨10.1007/978-3-030-05716-9_40⟩. ⟨hal-01896061⟩