Structural-RNN: Deep Learning on Spatio-Temporal Graphs
This addresses the problem of modeling structured sequences in computer vision, offering a generic framework that could empower new approaches, though it appears incremental in integrating existing tools.
The paper tackles the lack of high-level spatio-temporal structure in deep recurrent neural networks by proposing a method to combine spatio-temporal graphs with RNNs, resulting in a scalable, differentiable approach that shows large-margin improvements over state-of-the-art on tasks like human motion modeling.
Deep Recurrent Neural Network architectures, though remarkably capable at modeling sequences, lack an intuitive high-level spatio-temporal structure. That is while many problems in computer vision inherently have an underlying high-level structure and can benefit from it. Spatio-temporal graphs are a popular tool for imposing such high-level intuitions in the formulation of real world problems. In this paper, we propose an approach for combining the power of high-level spatio-temporal graphs and sequence learning success of Recurrent Neural Networks~(RNNs). We develop a scalable method for casting an arbitrary spatio-temporal graph as a rich RNN mixture that is feedforward, fully differentiable, and jointly trainable. The proposed method is generic and principled as it can be used for transforming any spatio-temporal graph through employing a certain set of well defined steps. The evaluations of the proposed approach on a diverse set of problems, ranging from modeling human motion to object interactions, shows improvement over the state-of-the-art with a large margin. We expect this method to empower new approaches to problem formulation through high-level spatio-temporal graphs and Recurrent Neural Networks.