Causal Intervention Learning for Multi-person Pose Estimation

Most of learning targets for multi-person pose estimation are based on the likelihood P(Y vertical bar X). However, if we construct the causal assumption for keypoints, named a Structure Causal Model (SCM) for the causality, P(Y vertical bar X) will introduce the bias via spurious correlations in the SCM. In practice, it appears as that networks may make biased decisions in the dense area of keypoints. Therefore, we propose a novel learning method, named Causal Intervention pose Network (ClposeNet). Causal intervention is a learning method towards solving bias in the SCM of keypoints. Specifically, under the consideration of causal inference, ClposeNet is developed based on the backdoor adjustment and the learning target will change into causal intervention P(Y vertical bar do(X)) instead of the likelihood P(Y vertical bar X). The experiments conducted on multi-person datasets show that ClposeNet indeed releases bias in the networks.