In this work, we present a 3D pose estimation model using a Transformer encoder based architecture with a trainable gate, PoseGate-Former.
確定! 回上一頁