THORN: Temporal Human-Object Relation Network for Action Recognition
This addresses action recognition for video analysis, particularly in first-person and human-object interaction contexts, with incremental improvements over existing methods.
The paper tackles action recognition by modeling human-object and object-object interactions, proposing THORN, an end-to-end network built on a 3D backbone, which achieves state-of-the-art performance on EPIC-Kitchen55 and EGTEA Gaze+ datasets.
Most action recognition models treat human activities as unitary events. However, human activities often follow a certain hierarchy. In fact, many human activities are compositional. Also, these actions are mostly human-object interactions. In this paper we propose to recognize human action by leveraging the set of interactions that define an action. In this work, we present an end-to-end network: THORN, that can leverage important human-object and object-object interactions to predict actions. This model is built on top of a 3D backbone network. The key components of our model are: 1) An object representation filter for modeling object. 2) An object relation reasoning module to capture object relations. 3) A classification layer to predict the action labels. To show the robustness of THORN, we evaluate it on EPIC-Kitchen55 and EGTEA Gaze+, two of the largest and most challenging first-person and human-object interaction datasets. THORN achieves state-of-the-art performance on both datasets.