1 paper
James Charles, Tomas Pfister, Derek Magee +2
We propose a personalized ConvNet pose estimator that automatically adapts itself to the uniqueness of a person's appearance to improve pose estimation in long videos. We make the…