Appearance-and-Relation Networks for Video Classification
This work addresses video action recognition, a fundamental problem in computer vision, by proposing a novel architecture that improves over existing methods, though it is incremental in its approach to spatiotemporal modeling.
The paper tackles spatiotemporal feature learning for video classification by introducing the Appearance-and-Relation Network (ARTNet), which uses SMART blocks to separately model appearance and relation, achieving superior performance on Kinetics, UCF101, and HMDB51 datasets compared to state-of-the-art methods.
Spatiotemporal feature learning in videos is a fundamental problem in computer vision. This paper presents a new architecture, termed as Appearance-and-Relation Network (ARTNet), to learn video representation in an end-to-end manner. ARTNets are constructed by stacking multiple generic building blocks, called as SMART, whose goal is to simultaneously model appearance and relation from RGB input in a separate and explicit manner. Specifically, SMART blocks decouple the spatiotemporal learning module into an appearance branch for spatial modeling and a relation branch for temporal modeling. The appearance branch is implemented based on the linear combination of pixels or filter responses in each frame, while the relation branch is designed based on the multiplicative interactions between pixels or filter responses across multiple frames. We perform experiments on three action recognition benchmarks: Kinetics, UCF101, and HMDB51, demonstrating that SMART blocks obtain an evident improvement over 3D convolutions for spatiotemporal feature learning. Under the same training setting, ARTNets achieve superior performance on these three datasets to the existing state-of-the-art methods.