Two-person Graph Convolutional Network for Skeleton-based Human Interaction Recognition
This addresses the challenge of effectively modeling spatial relationships in human-human interactions for computer vision applications, representing an incremental advance over existing graph convolutional network methods.
The paper tackled the problem of skeleton-based human interaction recognition by introducing a unified two-person graph to model inter-body and intra-body correlations, achieving state-of-the-art results on benchmarks like SBU and NTU-RGB+D datasets with accuracy improvements.
Graph convolutional networks (GCNs) have been the predominant methods in skeleton-based human action recognition, including human-human interaction recognition. However, when dealing with interaction sequences, current GCN-based methods simply split the two-person skeleton into two discrete graphs and perform graph convolution separately as done for single-person action classification. Such operations ignore rich interactive information and hinder effective spatial inter-body relationship modeling. To overcome the above shortcoming, we introduce a novel unified two-person graph to represent inter-body and intra-body correlations between joints. Experiments show accuracy improvements in recognizing both interactions and individual actions when utilizing the proposed two-person graph topology. In addition, We design several graph labeling strategies to supervise the model to learn discriminant spatial-temporal interactive features. Finally, we propose a two-person graph convolutional network (2P-GCN). Our model achieves state-of-the-art results on four benchmarks of three interaction datasets: SBU, interaction subsets of NTU-RGB+D and NTU-RGB+D 120.